In plain terms

The claim being assessed: The living human brain harbors a resident microbial community that contributes causally to neurodegeneration.

Where the evidence stands: Α — Speculative. An idea with a story behind it, but no evidence yet that it actually happens.

What this mechanism does: The evidence reviewed does not show this mechanism causing the outcome. That does not show it has no effect.

Strongest evidence for it: Preregistered primary-subclaim aggregation returns Α because SC1 living-human community existence remains unresolved under the low-biomass contamination gate. SC2 Β and SC3 Γ do not replace that prerequisite. Weakest-link and scope-matched sensitivities also return Α. Positive fish ecology and AD autopsy culture are explicitly retained rather than dismissed.

What decisive evidence would look like

In plain terms, researchers would change the proposed cause, confirm it changed as intended, and measure a meaningful health outcome while checking other explanations. The result should be able to support or count against the claim.

Relative explanatory fit: Alternative ranked ahead

What this review means

Some infections can affect the brain, and microbial material can be detected in tissue. This review asks the different question of whether healthy brain tissue normally contains a living resident microbial community that drives neurodegeneration.

Implications for health and treatment: Preregistered primary-subclaim aggregation returns Α because SC1 living-human community existence remains unresolved under the low-biomass contamination gate. SC2 Β and SC3 Γ do not replace that prerequisite. Weakest-link and scope-matched sensitivities also return Α. Positive fish ecology and AD autopsy culture are explicitly retained rather than dismissed.

Limits for individual care: 59 primary reports with 9 selected partial-body examinations; density, heterogeneous controls and incomplete access set the confidence, not correctness. The grade applies to the exact registered claim; stronger findings for narrower subclaims remain separate. No complete-paper audit or blinded calibration is claimed. This educational review does not establish an individual treatment recommendation.

Medical information notice: This review provides general health information, not medical advice. It cannot diagnose a condition or replace an individual discussion with a qualified clinician.

Source checking used complete abstracts/status records and selected original partial-body passages. No complete-paper audit or independent scientific peer review is claimed. Detailed access counts and limitations appear in the review.

A low grade is not an accusation. It means the causal work has not been done yet, which is the ordinary state of most actively studied mechanisms in biology.

At a glance

Sub-claims assessed separately

Figures

Explanatory figures summarizing this review. They carry no evidential weight of their own; where a figure and the review differ, the review governs.

Full assessment

From here the review is written for a technical reader: caps and thresholds, competing models, the independence basis, and the citation ledger.

Sources: 59 primary reports; claims: 63; verified: 63; unverified: 0. Verification is bounded attribution. Access: 9 partial-body and 50 complete abstract/status examinations. Literature cutoff: 2026-10-08; source examination UTC date: 2026-10-08. No complete-paper audit or blinded calibration claimed. Adversarial review: Generated by Sol 6.1; adversarially reviewed by Claude Opus 5.5 (Anthropic) on 2026-10-14. 4 corrections applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: Approved as revised.

AT A GLANCE

Human brain microbial molecules and occasional viable pathogens are credible possibilities; a reproducible resident community in living human parenchyma, causally driving neurodegeneration, is not established.

Deciding experiment: A prospectively registered multi-laboratory living-tissue study coupling quantitative strain-resolved genomes/RNA, recoverable viability and in situ localization to matched controls, followed by selective removal/reconstitution in a biologically matched model only after existence is demonstrated.

Belief versus evidence: Preregistered primary-subclaim aggregation returns Α because SC1 living-human community existence remains unresolved under the low-biomass contamination gate. SC2 Β and SC3 Γ do not replace that prerequisite. Weakest-link and scope-matched sensitivities also return Α. Positive fish ecology and AD autopsy culture are explicitly retained rather than dismissed.

Best-supported rival: Low-biomass reagent/host artifact. Randomize aliquots from promptly collected living human surgical parenchyma across independent laboratories and extraction lots, with vascular, tool, kit, PCR, RNA, culture and spike-in controls. Joint strain-linked viability/localization beats a taxonomic signature.

COMPETING MECHANISTIC MODELS

Rank 1: Low-biomass reagent/host artifact (Ε). Several technical sources can generate reproducible apparent bacterial profiles. Host-off-target disease taxa, dilution/batch interventions and cross-contamination benchmarks: PMIDs 33771222, 25387460, 31164452, 31239396. Positive surgical RNA/ISH, culture and edge/core patterns are not all explained away; method comparators are not universal brain nulls.

Rank 2: Disease-related transient microbial ingress (Ε). Pathogens or their molecules reach brain in selected disease settings without a stable community. Transient sepsis bacteria, HIV microbial molecules, AD Chlamydia isolates and installed P. gingivalis entry: PMIDs 29232157, 36202168, 18829386, 36631446. Cannot explain all purported homeostatic signals or establish common sporadic neurodegeneration causality.

Rank 3: Teleost-restricted residency (Ε). Healthy salmonid brain can contain viable bacteria under the reported conditions. Culture, perfusion, localization, negative controls and tagged colonization: PMID39292785. Shared study/fish sources, limited long-term persistence and no human replication prevent broad extrapolation.

Rank 4: Agonal/postmortem translocation (Δ). Cadaver communities and handling change tissue microbial signals. Cadaver interval studies and two-cohort brain processing effects: PMIDs 25091187, 33363519, 32310171. Long-delay cadavers cannot explain rapid surgical specimens by themselves.

Rank 5: Resident human brain community (Α). A viable stable brain community contributes to neurodegeneration. Surgical/autopsy RNA/ISH, positive AD culture and controlled taxonomic patterns: PMIDs 23355888, 18829386, 40284643. No contamination-resistant independent viable living-human community replication or strain-specific disease removal/rescue.

The tournament does not alter the nominal level. Compatible pathways and flexible alternatives are not automatically proved by explaining observations.

FULL REVIEW

Registered claim and scope

The living human brain harbors a resident microbial community that contributes causally to neurodegeneration.

Existence of a resident human brain community is the index dominant gate. Separate endogenous residence from infection, transient entry, postmortem translocation and contamination. Downstream disease causality cannot establish residence. Animal/teleost results remain bounded comparators.

Selection: public-interest rank 32 from the editorial ranking compiled 2026-09-23; categorical prior verdicts were visible. Private fresh review; earlier documents are preserved.

Evidence and causal interpretation

The central distinction is residence in living human parenchyma, rather than detection of microbial material. Human products, diseased autopsy isolates, transient sepsis dissemination and viable fish bacteria are different propositions. A viable Chlamydia isolate from each of two AD brains is substantial positive evidence for organisms in those specimens (PMID18829386), but neither healthy living community nor disease necessity follows.

The 2013 original includes epilepsy resections and autopsy material, with RNA/ISH, glial peptidoglycan and nested water/tool controls (PMID23355888). Its Rag1-null mouse transfer used unheated ODC14 and heated ODC13, so donor differences remain a confound; RNA persisted seven weeks without neurological deficits or increased measured immune transcripts. The transfer therefore does not establish neurodegenerative phenotype causation.

The 2021 host-aware study examined 47 postmortem and three surgical brains, multiple extraction/library blanks and mock dilutions (PMID33771222). Disease-differentiating apparent Clostridia were host off-targets. Human copy loads were not literally zero and one candidate taxon persisted at low levels, consistent with contamination. The mock missed Listeria and under-recovered Bacillus without mechanical lysis, so the conclusion is an assay/cohort null, not proof of universal sterility.

AD molecular positives deserve explicit evaluation. Emery 2017 normalized and reamplified libraries, acknowledges human 18S coamplification, and did not design read numbers as absolute-load measures (PMID28676754). Alonso 2018 reports multiple PCR negatives and sterile rapid-autopsy handling (PMID29881346); Consensus could not retrieve those negative-control details. That provider limitation is not study absence, while sequenced independent bacterial extraction blanks and viability remain unverified in the selected source passages.

AlzPI work strengthens sampling/analysis without resolving residence. The 2023 study has 130 samples from 32 subjects and four blanks; its LDA temporal order assumes health precedes disease rather than observing progression (PMID37780846). The 2025 edge/core study has 86 samples from 31 specimens and four extraction blanks; lack of edge enrichment argues against one surface-contamination pattern but is not equivalence or exclusion of reagent, vascular and intratissue preanalytic artifacts (PMID40284643). Multiple samples per donor are not independent participants.

Salmonid evidence is substantially stronger for its restricted ecology: perfusion, localized organisms, RNA/DNA concordance, 54 cultured brain isolates, culture-buffer/CSF negatives and nonmatching environmental taxa (PMID39292785). Tagged gut Plesiomonas enters brain and declines by day 14, with persistence beyond that explicitly unresolved. Genome comparisons use isolates from two fish. Adult Chinook LPS/amyloid proximity does not show a microbial neurodegenerative cause.

Contamination is itself experimentally demonstrated: pure-culture dilutions across kits/facilities create apparent communities and false age clusters (PMID25387460), while blank-only filtering can delete true organisms and miss contaminants (PMID31164452). Plate cross-talk, host mapping, load and stochastic-amplification reports add complementary methods evidence. They inform brain controls but do not count as direct demonstrations of brain absence.

Sepsis transient colonization and HIV microbial-molecule ingress support disease-related entry (PMIDs 29232157, 36202168). P. gingivalis inoculation and repeated bacteremia alter disease/barrier readouts (PMIDs 30746447, 36631446). These are plausible contributory pathways with installed exposure or barrier dysfunction, not proof of the registered resident-human community. The preregistered existence gate remains Α.

HYPOTHESIS STATE EVALUATION FRAMEWORK (HSEF v5.3)

Subclaim matrix and aggregation

Preregistered aggregation: primary-subclaim. Preregistered primary-subclaim aggregation returns Α because SC1 living-human community existence remains unresolved under the low-biomass contamination gate. SC2 Β and SC3 Γ do not replace that prerequisite. Weakest-link and scope-matched sensitivities also return Α. Positive fish ecology and AD autopsy culture are explicitly retained rather than dismissed. Sensitivities: weakest-link Α; primary-subclaim Α; scope-matched Α.

Decisive-perturbation inventory and causal tally

[1] Surgical epilepsy and autopsy brain show bacterial RNA, glial antigen/ISH and seven-week RNA recovery after implantation in Rag1-null mice, with PCR water and tool controls. No human culture or resident-community disease phenotype; heated ODC13 and unheated ODC14 are different donors, confounding the viability contrast. SURG is epilepsy tissue, not healthy population sampling.

[2] 47 postmortem and three surgical samples were analyzed with extraction/PCR blanks, mock dilutions, qPCR and host alignment; disease-differentiating taxa were host off-targets. Null is bounded to this assay/cohort. Positive mock missed Listeria and under-recovered Bacillus without mechanical lysis; residual human bacterial load was not literally zero.

[5] 86 samples from 31 specimens, 14 AD/17 controls, with four extraction blanks and edge/core sampling; no edge enrichment and some disease-associated taxa after decontamination. No edge enrichment is not statistical equivalence or absence of all contamination; both edge and core may share reagent/intravascular sources. No culture or prospective disease ordering.

[22] Sepsis mice have viable brain bacteria around day five but not day 14; human postmortem sepsis brain signals correlate with the systemic source. Demonstrates transient disease-associated dissemination, not homeostatic residency or sporadic neurodegenerative causality.

[23] Perfused trout brains show RNA/DNA, localized bacteria and 54 cultured isolates; buffer/CSF negatives, environment controls, antibiotics and tagged strain colonization corroborate fish residence. Not human evidence. Tagged-strain persistence declines by day 14, brain isolates come from two fish for genome comparisons, and adult salmon LPS/amyloid association is not a causal degeneration intervention.

[24] 23 healthy pregnant women, 116 CSF/control samples and paired metatranscriptomics show no active viable microbial community after contamination correction. CSF is a distinct compartment and pregnancy a restricted population; cannot prove universal absence of parenchymal organisms.

[27] Two AD cohorts show processing/postmortem effects greater than the limited AD-control differences. DNA may still be present; the null concerns disease signatures and study bias, not an absolute sterility proof.

[28] 22 HIV/11 control human and 11 SIV/4 control primate specimens show microbial molecules; peptidoglycan alters CCR5/HIV and neuronal mitochondrial readouts. Translocated microbial molecules can act without resident viable bacteria; HIV-linked immune/barrier pathology is not sporadic neurodegeneration. Online 2022, issue 2023.

[29] Serial pure-culture dilutions across three facilities and multiple kits produce kit-specific false profiles; cohort-age clustering tracked kit batches. Demonstrates a causal artifact mechanism, not proof that every positive human brain signal is contaminated; tested quantitative floors are protocol-specific.

[30] 144 negative controls over five years show kit/laboratory/researcher/month variation in contaminant communities. A historical contaminant list cannot classify every future tissue taxon; this is methods evidence, not direct human-brain absence.

[31] Low-biomass metagenomic decontamination uses coverage and batch variation and is tested in clinical cfDNA settings. Algorithmic benchmarking in urine/amniotic settings does not measure brain viability or guarantee filtering accuracy there.

[32] 23-subject airway study attributes substantial low-biomass profiles to laboratory contamination. Airway-specific fractions do not estimate the brain fraction; matched brain recovery controls remain necessary.

[33] A contaminant-controlled workflow detects microbial cfDNA exceeding blanks in plasma samples. Circulating DNA can exist without live resident plasma organisms and may contribute to brain tissue signals.

[34] Mock serial dilutions and clinical respiratory samples reveal stochastic profiles below the tested bacterial-load range. The numerical copy threshold is assay-specific and cannot be imposed as a universal brain cutoff.

[35] Seven blank libraries contain 14 viral contaminants, demonstrating reagent-derived RNA/viral reads. RNA detection alone cannot establish viability, but contamination in these reagents is not a universal explanation of all viral tissue findings.

[36] Synthetic/public benchmarks assess host-removal accuracy; long-read performance remains incomplete. A host-removal tool is not direct biological evidence for or against brain residence.

[37] Pan-cancer profiles use center/batch and matched tissue information to distinguish candidate tissue signals from contaminants. Cancer/mucosal tissue validation does not establish brain microbiota or living strain identity.

[38] CS-SCORE benchmarks human-read removal with computational efficiency benefits. Speed and genome filtering do not certify low-copy bacterial identification, tissue localization or viability.

[39] Squeegee cross-sample contaminant identification is benchmarked against control-based approaches. Data-driven decontamination does not replace independent physical blanks or prove brain community existence.

[40] Defined mammalian/bacterial FFPE blocks compare recovery with qPCR, 16S and whole-genome methods. Useful assay recovery standard; fixed mock blocks are not viable organisms in human brain.

[41] decontam uses frequency/prevalence with culture dilution and biological cohorts to identify contaminants. Model assumptions can fail with cross-contamination or shared true/background taxa; filtered signal is not automatically true residence.

[42] Known mock dilutions quantify contamination and false deletion; blank-only filtering removes 20.1–33.9% of true ASVs and misses many contaminants. No algorithm is perfect across mixtures. Unknown-source SourceTracker fails substantially, while known-source success cannot be assumed for brain.

[43] qPCR bacterial-load-aware removal preserves known culture members and reduces tested contaminants. Some sample cross-contaminants escape blank-based correction; recovery must be tested for the actual brain protocol.

[44] Alternative two-stage 16S/ITS amplification improves low-biomass sequencing in mocks and clinical samples. Amplification yield or consistency cannot demonstrate microbial viability or absence of stochastic errors below recovery limits.

[45] 54-species mock and environmental microsamples validate very-low-input shotgun libraries with explicit complexity/contamination limits. Real microspatial variation is possible; low-input validation in known communities is not automatic brain validity.

[46] Unique bacterial assignments across two laboratories reveal well-to-well contamination, strongest at low biomass. Cross-contamination can introduce real tissue taxa into blanks; blanket deletion of blank taxa can therefore erase true organisms.

[47] Mock and bile analyses with repeated controls show stochastic amplification and incomplete single-blank coverage. Clinical bile is not brain; quantitative per-sample decision rules require protocol-specific recovery rather than taxon blacklists.

[48] Six inference approaches across biomass dilutions compare false sequence/OTU calls; ASV methods improve precision. Sequence inference does not by itself distinguish true low-load tissue biology from contamination.

[49] DNase treatment of PCR master mixes reduces measured reagent contamination about 99% in the tested protocol. Kit extraction is not always the largest contaminant source; treatment must preserve assay sensitivity. Online 2018, issue 2019.

[50] 79 samples from 54 patients include 51 intracranial abscesses, with sequencing detecting broader polymicrobial pathogens. Diagnostic infection is a positive assay control, not evidence of a normal resident human brain community.

[51] Nanopore 16S of 27 abscess pus samples detects more anaerobic/polymicrobial findings than culture. Abscess load, inflammation and collection differ from healthy low-biomass brain; detection is not homeostasis.

[52] Prospective 20-patient abscess comparison identifies 72 sequenced strains versus 22 culture strains and expands species spectrum. Clinically localized infection establishes microbes can enter brain, without normal residence or neurodegenerative community causality.

[53] 63 abscess cases yield 18 ISH positives; reported 35% sensitivity/93% specificity versus the composite comparator. Negative ISH cannot exclude all infection, and abscess diagnostic performance cannot establish healthy-tissue assay sensitivity. Online 2024, issue 2025.

[54] Two AD brain-derived Chlamydia pneumoniae isolates were cultured through HEp-2 passages and characterized genetically. A substantive viability positive in diseased autopsy tissue; does not establish community residence in living brains, healthy controls or AD necessity.

[56] Human AD gingipain signals accompany inoculated-mouse bacterial load and small-molecule inhibition effects on neurodegenerative readouts. Forced pathogen exposure and protease inhibition are not removal/rescue of an endogenous human resident community.

[57] Repeated intravenous P. gingivalis exposure changes BBB transcytosis; Cav/Mfsd-related interventions affect barrier readouts. Installed bacteremia can permit entry; it cannot establish stable homeostatic human residence or sporadic-disease primacy.

[58] 38-woman placenta culture/qPCR and selected sequencing distinguish surface organisms from internal signals matching blanks/database artifacts. Sterile-tissue comparator highlights contamination control, but placenta absence does not prove adult brain absence.

[59] Cesarean fetal brain/placenta signals do not exceed controls; isolated C. acnes also occurs in controls while positive controls work. Restricted fetal macaque setting cannot settle adult human brain ecology; a valuable organ/method comparator only.

Tally: 38 discriminating contrasts. Null and negative contrasts are discriminators, not positive threshold crossings. Reports and replications are not independent causal lines. Shared cohorts, experimental platforms, transgenic lineages and repeated outcomes do not constitute fully independent causal lines.

Cap audit, precedence and Β/Γ/Δ boundary

The 04 low-biomass Α existence gate governs SC1. Forced transfer can cross Δ only for a demonstrated recipient phenotype; the 2013 transfer had no neurological disease phenotype and different donors in its heat contrast. Salmon culture supports its own restricted ecological model but cannot lift living-human residence. Convergent human positives are assessed together; their shared unexcluded viability/preanalytic failure modes still leave the gate unresolved.

Gamma hard-rule audit: No endogenous human resident-community strain subtraction/addition with disease phenotype has been established. Forced human-homogenate or pathogen transfer is bounded to the recipient system; antigen, DNA and RNA signals do not alone establish viability.

SC1 cap: “Use for ideas sustained mainly by contaminated signal, weak association, or conceptual enthusiasm.” Α: orthogonal positive surgical/autopsy signals exist, but no independent viable strain-resolved living-human community replication excludes the demonstrated artifact alternatives. This is an unresolved existence gate, not a finding of universal sterility.

SC2 cap: “Association-only microbiome work caps at Β.” Β: controls and edge/core comparisons improve specificity, but no complete human case rules out reagent, host, vascular and preanalytic origins together; host-aware null and assay-recovery failures constrain both directions.

SC3 cap: “Association plus mechanistic in vitro support, or host-pathway plausibility, caps at Γ.” Γ: microbial products and installed infection can change relevant readouts, but resident-human community removal/rescue and neurodegenerative phenotype movement are absent from the examined corpus.

SC1 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. Α: orthogonal positive surgical/autopsy signals exist, but no independent viable strain-resolved living-human community replication excludes the demonstrated artifact alternatives. This is an unresolved existence gate, not a finding of universal sterility.

SC2 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. Β: controls and edge/core comparisons improve specificity, but no complete human case rules out reagent, host, vascular and preanalytic origins together; host-aware null and assay-recovery failures constrain both directions.

SC3 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. Γ: microbial products and installed infection can change relevant readouts, but resident-human community removal/rescue and neurodegenerative phenotype movement are absent from the examined corpus.

Upward threshold ladder

SC1 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. A prespecified causal test matching this subclaim, with mechanistic engagement, meaningful phenotype change and an appropriate negative control.

SC2 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. A prespecified causal test matching this subclaim, with mechanistic engagement, meaningful phenotype change and an appropriate negative control.

SC3 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. A prespecified causal test matching this subclaim, with mechanistic engagement, meaningful phenotype change and an appropriate negative control.

Upper ladder: Strain-resolved living-human colonization, selective removal/rescue, dose-response and eradication natural experiments are live upper-ladder routes. None in the examined corpus establishes the entire resident-human community claim; ethical limits on healthy sampling do not constitute disproof.

Independence: 2013 RNA/ISH/transfer are a shared sample platform; recurrent fungal papers share authors/assays; 2023/2025 AlzPI material may overlap. Fish sampling across species is within one investigative platform. Primary-report counts are not independent human biological replications.

Different error exclusions, not equivalent biological replication [2, 29, 42, 46]: Host mispriming, reagent lots and plate cross-talk are distinct tested error mechanisms.

Falsification and pitfalls

Host-aware analysis removes apparent PD-specific bacterial differences. PMID33771222 manual alignment identifies disease-differentiating taxa as human off-targets. Additional downgrade: False.

RNA or antigens equated with viability. Human molecular positives versus culture; reagent RNA is possible.

Absent provider excerpt treated as absent study control. Alonso 2018 documents multiple PCR negatives, though sequenced bacterial extraction blanks remain unverified.

Nonsignificance treated as equivalence. 2025 edge/core posterior intervals include zero but do not set an equivalence margin.

Inferred LDA order treated as prospective disease time. 2023 source explicitly uses a health-before-disease ordering assumption.

Fish, infection and fluid results transferred to living human parenchyma. Scope boundaries in SC1/SC3.

Adversarial steelman: above, then below

Steelman Β: independent molecular methods, surgical samples, rapid autopsy and cultured AD isolates support some genuine microbial material and occasional organisms. This justifies investigation, but does not establish the stable living-human community required by SC1.

There is no rung below Α. An assertion of refuted or universally sterile would overread the assay-specific null, culture positives and strong fish evidence.

Keep Α as insufficient establishment of the registered existence gate, without converting it into universal absence; retain narrow positive findings and higher restricted-model scores.

Final score, causal direction and M/S/D

Preregistered primary-subclaim aggregation returns Α because SC1 living-human community existence remains unresolved under the low-biomass contamination gate. SC2 Β and SC3 Γ do not replace that prerequisite. Weakest-link and scope-matched sensitivities also return Α. Positive fish ecology and AD autopsy culture are explicitly retained rather than dismissed.

Canonical step 2 applies: no causal dependency in either direction is demonstrated for the nominal resident-human community. This label records co-occurrence without demonstrated dependency, not proof of biological inertness. Product ingress and installed infection can have causal effects in separate settings.

Momentum: Mechanistic only. Active mechanistic and methodological work only. The named artifacts — 2023 full-length profiling, 2024 salmon culture and 2025 human edge/core sampling — are completed publications, and the review itself states that no current recruiting trial is asserted; a completed report is not an experiment being run that could cross the next threshold. No ongoing multi-laboratory living-human viability study addressing the SC1 existence gate was identified.

Stability: Stable. No named pending result is asserted to be likely to move the registered score imminently.

Divergence: Contested; no single implied-level arithmetic. These dimensions do not change level or confidence.

PROPOSED DECISIVE EXPERIMENTS

A prospectively registered multi-laboratory living-tissue study coupling quantitative strain-resolved genomes/RNA, recoverable viability and in situ localization to matched controls, followed by selective removal/reconstitution in a biologically matched model only after existence is demonstrated.

Readout: Absolute organism recovery above validated protocol background, viable strain concordance, vascular exclusion and separate neuronal/functional endpoints. Controls: Independent collection/tool/reagent/PCR blanks, host-aware mapping, blind spike-ins with difficult-to-lyse taxa, matched culture controls and same-donor heat comparisons. Supporting: The same viable localized strains reproducibly exceed controls across laboratories and selective removal/rescue moves neurodegenerative phenotype. Falsifying: Signal follows lots, host sequences, blood or collection delay, or confirmed removal leaves disease intact.

RESEARCH TIMELINE

KNOWLEDGE GAPS

Independent quantitative viable-community replication in promptly collected living human parenchyma is missing.

Vascular content, peripheral products and tissue localization must be distinguished by strain-matched orthogonal assays.

Protocol-specific detection/recovery limits must include hard-to-lyse organisms; the 2021 positive-control misses matter.

Use same-donor aliquots for heated/unheated transfer; the 2013 donor contrast cannot isolate viability.

Separate RNA contamination, dormant/unculturable possibilities and recoverable viability without defining residence solely from whichever assay is positive.

Cross-sectional microbial class ordering needs longitudinal or intervention confirmation; no inferred timeline should be presented as observed disease progression.

A resident-community disease experiment requires selective subtraction/addition with strain engagement and meaningful neuronal endpoints, distinct from forced infection.

Fish persistence beyond 14 days and its functional consequences remain open, with no direct human extrapolation.

Edge/core nonsignificance needs predefined equivalence margins, absolute quantification and independent collection controls.

METHODS AND CONSENSUS CONTRIBUTION

Consensus was used through the signed-in browser for discovery, counterevidence, research-gap and timeline suggestions, and attempted checks of specific indexed full-text passages. Returned Methods/Results passages were reconciled with original-paper passages where available; failed retrievals, source substitutions and unverified aspects are recorded in the case reconciliation. Successful passage checks are distinguished from provider summaries. Provider screening counts, citation counts and meters were not used as primary counts or HSEF scores. This source log records access depth; it does not claim complete-paper or supplement review. The monthly API allowance was conserved: zero Consensus API calls in this batch.

REVIEWER NOTES

59 primary reports examined, including explicitly identified methods, diagnostic infection and organ comparators; nine selected original-body examinations. One read protocol chapter (PMID33161546) excluded from primary count. Consensus two Pro browser queries, zero API/Deep; targeted output supplies useful provenance/controls but cannot certify unavailable excerpts. All body credit comes from independently recorded public-original paragraphs. No complete-paper/supplement or blinded calibration claim.

Search leads, reviews, protocols and correction notices do not pad the primary count. Downloaded XML does not earn examination credit. Linked notices remain attached; unexamined correction details limit inference. This is a coordinator source/claim consistency assessment, not independent blinded calibration.

ADVERSARIAL REVIEW AND APPROVAL

Authoring model: Sol 6.1. Adversarial reviewer: Claude Opus 5.5 (Anthropic). Date: 2026-10-14. Edition reviewed: the 9 October 2026 revised edition under consensus_consistency_audit_2026-10-09/revised/, compared line-by-line against the earlier edition in hsef_v53_popularity_ranks_31_35_2026-10-08/.

Scope of checks. All 59 REFERENCES PMIDs verified against PubMed metadata (59/59 resolve to the stated title, first author, journal and year; no retraction or expression-of-concern notice attaches to any source), plus the excluded protocol record 33161546. Because contamination control is the crux of this claim, the review's treatment of negative and blank controls and of the contamination counter-explanation was checked against the retained original full texts rather than abstracts alone: PMIDs 40284643, 33771222, 23355888, 28676754, 29881346, 39292785 were re-read in full-text/*.sections.json, and PMIDs 18829386, 39061418, 39509244, 25387460, 31164452, 31239396 against their abstracts. Scoring logic checked against the frozen framework: hard caps before aggregation, the declared primary-subclaim aggregation, Ζ independence basis, Β/Γ/Δ boundary, confidence-band text under lead policy, role-of-mechanism derivation order, the Momentum artifact rule, and the known upstream builder defects.

Verification of the contamination-control treatment (the crux). Every control claim checked matched the cited original, and in each case the review's reading was the conservative one:

• PMID 40284643 (AlzPI edge/core, 2025). The review's "86 samples from 31 specimens, 14 AD/17 controls, four extraction blanks" matches the paper's Methods verbatim, including "Four blank extraction controls … without sample input" and the use of decontam plus microDecon. The paper's own abstract says "30 brain specimens" while its Methods say 31 (14 + 17); the review follows the Methods, which is correct. The review also correctly declines the paper's title claim ("Absence of Postmortem Contamination"): no significant edge-versus-core difference is not an equivalence test and does not exclude reagent, vascular or intratissue preanalytic sources, and the review says exactly that.
• PMID 33771222 (host-aware null, 2021). "47 postmortem and three surgical brains" matches (N = 47 postmortem, PD 25 / HC 22; n = 3 fresh cortex from living volunteers at resective surgery). The positive-control limitation is reported accurately: the mock community recovered only six of eight expected species, Listeria monocytogenes was absent and Bacillus subtilis under-recovered, with no mechanical lysis performed. The review uses this to bound the null in both directions rather than to dismiss it.
• PMID 23355888 (2013 surgical/autopsy RNA and transfer). The heated/unheated assignment is correct — the originals state that ODC13 and ODC14 homogenates were implanted with and without prior heat treatment respectively — so the review's donor-confound objection is a real design limitation, not an invention; the seven-week RNA recovery without neurological deficit is also as reported.
• PMID 28676754 (Emery 2017). The paper itself states it "was carried out with normalized, re-amplified libraries and therefore was not designed to assess actual bacterial numbers", and discusses human 18S co-amplification. The review reproduces both limits.
• PMID 29881346 (Alonso 2018). "Multiple PCR negatives and sterile rapid-autopsy handling" is supported: separate rooms and glassware, aliquoted reagents, positive-displacement pipettes, aerosol-resistant tips and multiple negative controls; sterile instruments in a laminar flow cabinet; rapid neuropathological autopsy with a mean postmortem interval of 4.5 h. The review's hedge that sequenced bacterial extraction blanks remain unverified in the accessed passages is accurate — the retained passages document PCR negative controls, not sequenced blank libraries.
• PMID 39292785 (salmonid residency). Perfusion with haemoglobin verification, lysis-buffer-only negative plates, CSF from the same animals plated identically, environmental taxa that did not match the fish isolates, 54 brain isolates, tagged Plesiomonas loads significantly lower at day 14 than at 3 hours, and comparative genomics on isolates "from two different laboratory rainbow trout" are all as the review states.

Rejected findings (checked, no change).

1. "Best-supported rival: Low-biomass reagent/host artifact" — checked against the score card: M2 has is_nominal: false and rank_in_tournament: 1, and the nominal model M1 ("Resident human brain community") ranks fifth. This review does not carry the renderer defect that prints the nominal model as the rival.
2. Primary-subclaim aggregation returning Α — SC1 is the registered existence gate and is scope-matched to the index claim; weakest-link and scope-matched sensitivities also return Α, and the Β/Γ subclaims are explicitly stated not to substitute for the prerequisite. No Ε subclaim is doing the work.
3. REVIEWER-NOTES contradiction — not present here. METHODS AND CONSENSUS CONTRIBUTION and REVIEWER NOTES both record browser Consensus use (two Pro queries, zero API/Deep) and state that body credit comes only from independently recorded original paragraphs, which is consistent with the spliced provenance section. Only the approval clause (finding 1) was false.
4. The 30-versus-31-specimen discrepancy in PMID 40284643 — the review's 31 follows the paper's Methods and is correct; the paper's own abstract is the inconsistent text. Recorded here rather than changed.
5. Confidence band Moderate paired with an Α level — the framework states explicitly that the two axes are independent and that a low level can carry a confident assessment; "we are confident the existence gate is not yet crossed" is the intended reading.
6. Role of mechanism "epiphenomenon" — derivation step 2 is correctly applied to the nominal resident-community proposition and explicitly not taken from tournament rank; the review separately preserves causal effects of product ingress and installed infection.
7. Stability "Stable" — no named pending trigger is asserted, which the framework says is stability by definition.
8. Collective-author rendering defect — no reference in this list has a consortium first author.
9. Headline counts — 59 sources, 63 claims, 63 verified, 9 partial-body plus 50 abstract/status (= 59) all agree between the banner, the score card and the reference list.
10. The tally of "38 discriminating contrasts" with no positive qualifying unit — consistent with the inventory as written; null and negative contrasts are correctly treated as discriminators rather than crossings, and the positive items (AD Chlamydia culture, salmonid residency) are retained at their own restricted scopes rather than discarded.

Residual limitations. Access is bounded: 9 partial-body and 50 abstract/status examinations. Six of the nine body-level sources were re-read here; the remaining three were not, so a small number of body statements rest on the original reviewer's reading alone. Whether sequenced bacterial extraction blanks exist for PMID 29881346 remains unresolved in the accessed passages, which the review states. The level and confidence of this revised edition were re-derived and are unchanged from the earlier edition; only Momentum, which modifies no letter, moved.

Approval: Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-14.

REFERENCES

1. Branton WG, Ellestad KK, Maingat F et al. Brain microbial populations in HIV/AIDS: α-proteobacteria predominate independent of host immune status. PloS one. 2013. PMID 23355888. DOI: 10.1371/journal.pone.0054673. Access: full_text_partial. Study: PMID23355888.

2. Bedarf JR, Beraza N, Khazneh H et al. Much ado about nothing? Off-target amplification can lead to false-positive bacterial brain microbiome detection in healthy and Parkinson's disease individuals. Microbiome. 2021. PMID 33771222. DOI: 10.1186/s40168-021-01012-1. Access: full_text_partial. Study: PMID33771222.

3. Emery DC, Shoemark DK, Batstone TE et al. 16S rRNA Next Generation Sequencing Analysis Shows Bacteria in Alzheimer's Post-Mortem Brain. Frontiers in aging neuroscience. 2017. PMID 28676754. DOI: 10.3389/fnagi.2017.00195. Access: full_text_partial. Study: PMID28676754.

4. Emery DC, Davies M, Cerajewska TL et al. High resolution 16S rRNA gene Next Generation Sequencing study of brain areas associated with Alzheimer's and Parkinson's disease. Frontiers in aging neuroscience. 2022. PMID 36570533. DOI: 10.3389/fnagi.2022.1026260. Access: abstract_metadata. Study: PMID36570533.

5. Thwe MN, Moné Y, Sen B et al. Microspatial Heterogeneities and the Absence of Postmortem Contamination in Alzheimer's Disease Brain Microbiota: An Alzheimer's Pathobiome Initiative (AlzPI) Study. Microorganisms. 2025. PMID 40284643. DOI: 10.3390/microorganisms13040807. Access: full_text_partial. Study: PMID40284643.

6. Moné Y, Earl JP, Król JE et al. Evidence supportive of a bacterial component in the etiology for Alzheimer's disease and for a temporal-spatial development of a pathogenic microbiome in the brain. Frontiers in cellular and infection microbiology. 2023. PMID 37780846. DOI: 10.3389/fcimb.2023.1123228. Access: full_text_partial. Study: PMID37780846.

7. Jones TB, Chu P, Wilkey B et al. Regional Differences in Microbial Infiltration of Brain Tissue from Alzheimer's Disease Patients and Control Individuals. Brain sciences. 2024. PMID 39061418. DOI: 10.3390/brainsci14070677. Access: abstract_metadata. Study: PMID39061418.

8. Alonso R, Pisa D, Carrasco L et al. Brain Microbiota in Huntington's Disease Patients. Frontiers in microbiology. 2019. PMID 31798558. DOI: 10.3389/fmicb.2019.02622. Access: abstract_metadata. Study: PMID31798558.

9. Pisa D, Alonso R, Rábano A et al. Different Brain Regions are Infected with Fungi in Alzheimer's Disease. Scientific reports. 2015. PMID 26468932. DOI: 10.1038/srep15015. Access: abstract_metadata. Study: PMID26468932.

10. Alonso R, Pisa D, Aguado B et al. Identification of Fungal Species in Brain Tissue from Alzheimer's Disease by Next-Generation Sequencing. Journal of Alzheimer's disease : JAD. 2017. PMID 28387676. DOI: 10.3233/JAD-170058. Access: abstract_metadata. Study: PMID28387676.

11. Pisa D, Alonso R, Fernández-Fernández AM et al. Polymicrobial Infections In Brain Tissue From Alzheimer's Disease Patients. Scientific reports. 2017. PMID 28717130. DOI: 10.1038/s41598-017-05903-y. Access: abstract_metadata. Study: PMID28717130.

12. Pisa D, Alonso R, Carrasco L et al. Parkinson's Disease: A Comprehensive Analysis of Fungi and Bacteria in Brain Tissue. International journal of biological sciences. 2020. PMID 32174790. DOI: 10.7150/ijbs.42257. Access: abstract_metadata. Study: PMID32174790.

13. Pisa D, Alonso R, Rábano A et al. Corpora Amylacea of Brain Tissue from Neurodegenerative Diseases Are Stained with Specific Antifungal Antibodies. Frontiers in neuroscience. 2016. PMID 27013948. DOI: 10.3389/fnins.2016.00086. Access: abstract_metadata. Study: PMID27013948.

14. Alonso R, Pisa D, Marina AI et al. Evidence for fungal infection in cerebrospinal fluid and brain tissue from patients with amyotrophic lateral sclerosis. International journal of biological sciences. 2015. PMID 25892962. DOI: 10.7150/ijbs.11084. Access: abstract_metadata. Study: PMID25892962.

15. Pisa D, Alonso R, Juarranz A et al. Direct visualization of fungal infection in brains from patients with Alzheimer's disease. Journal of Alzheimer's disease : JAD. 2015. PMID 25125470. DOI: 10.3233/JAD-141386. Access: abstract_metadata. Study: PMID25125470.

16. Pisa D, Alonso R, Rábano A et al. Fungal Enolase, β-Tubulin, and Chitin Are Detected in Brain Tissue from Alzheimer's Disease Patients. Frontiers in microbiology. 2016. PMID 27872620. DOI: 10.3389/fmicb.2016.01772. Access: abstract_metadata. Study: PMID27872620.

17. Alonso R, Pisa D, Fernández-Fernández AM et al. Infection of Fungi and Bacteria in Brain Tissue From Elderly Persons and Patients With Alzheimer's Disease. Frontiers in aging neuroscience. 2018. PMID 29881346. DOI: 10.3389/fnagi.2018.00159. Access: full_text_partial. Study: PMID29881346.

18. Pisa D, Alonso R, Marina AI et al. Human and Microbial Proteins From Corpora Amylacea of Alzheimer's Disease. Scientific reports. 2018. PMID 29959356. DOI: 10.1038/s41598-018-28231-1. Access: abstract_metadata. Study: PMID29959356.

19. Alonso R, Pisa D, Rábano A et al. Cerebrospinal Fluid from Alzheimer's Disease Patients Contains Fungal Proteins and DNA. Journal of Alzheimer's disease : JAD. 2015. PMID 26401766. DOI: 10.3233/JAD-150382. Access: abstract_metadata. Study: PMID26401766.

20. Alonso R, Pisa D, Marina AI et al. Fungal infection in patients with Alzheimer's disease. Journal of Alzheimer's disease : JAD. 2014. PMID 24614898. DOI: 10.3233/JAD-132681. Access: abstract_metadata. Study: PMID24614898.

21. Zhan X, Stamova B, Jin LW et al. Gram-negative bacterial molecules associate with Alzheimer disease pathology. Neurology. 2016. PMID 27784770. DOI: 10.1212/WNL.0000000000003391. Access: abstract_metadata. Study: PMID27784770.

22. Singer BH, Dickson RP, Denstaedt SJ et al. Bacterial Dissemination to the Brain in Sepsis. American journal of respiratory and critical care medicine. 2018. PMID 29232157. DOI: 10.1164/rccm.201708-1559OC. Access: abstract_metadata. Study: PMID29232157.

23. Mani A, Henn C, Couch C et al. A brain microbiome in salmonids at homeostasis. Science advances. 2024. PMID 39292785. DOI: 10.1126/sciadv.ado0277. Access: full_text_partial. Study: PMID39292785.

24. Kang Y, Ji X, Guo L et al. Cerebrospinal Fluid from Healthy Pregnant Women Does Not Harbor a Detectable Microbial Community. Microbiology spectrum. 2021. PMID 34908504. DOI: 10.1128/Spectrum.00769-21. Access: abstract_metadata. Study: PMID34908504.

25. Can I, Javan GT, Pozhitkov AE et al. Distinctive thanatomicrobiome signatures found in the blood and internal organs of humans. Journal of microbiological methods. 2014. PMID 25091187. DOI: 10.1016/j.mimet.2014.07.026. Access: abstract_metadata. Study: PMID25091187.

26. Lutz H, Vangelatos A, Gottel N et al. Effects of Extended Postmortem Interval on Microbial Communities in Organs of the Human Cadaver. Frontiers in microbiology. 2020. PMID 33363519. DOI: 10.3389/fmicb.2020.569630. Access: abstract_metadata. Study: PMID33363519.

27. Westfall S, Dinh DM, Pasinetti GM et al. Investigation of Potential Brain Microbiome in Alzheimer's Disease: Implications of Study Bias. Journal of Alzheimer's disease : JAD. 2020. PMID 32310171. DOI: 10.3233/JAD-191328. Access: abstract_metadata. Study: PMID32310171.

28. Branton WG, Fernandes JP, Mohammadzadeh N et al. Microbial molecule ingress promotes neuroinflammation and brain CCR5 expression in persons with HIV-associated neurocognitive disorders. Brain, behavior, and immunity. 2023. PMID 36202168. DOI: 10.1016/j.bbi.2022.09.019. Access: abstract_metadata. Study: PMID36202168.

29. Salter SJ, Cox MJ, Turek EM et al. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC biology. 2014. PMID 25387460. DOI: 10.1186/s12915-014-0087-z. Access: full_text_partial. Study: PMID25387460.

30. Weyrich LS, Farrer AG, Eisenhofer R et al. Laboratory contamination over time during low-biomass sample analysis. Molecular ecology resources. 2019. PMID 30887686. DOI: 10.1111/1755-0998.13011. Access: abstract_metadata. Study: PMID30887686.

31. Burnham P, Gomez-Lopez N, Heyang M et al. Separating the signal from the noise in metagenomic cell-free DNA sequencing. Microbiome. 2020. PMID 32046792. DOI: 10.1186/s40168-020-0793-4. Access: abstract_metadata. Study: PMID32046792.

32. Drengenes C, Wiker HG, Kalananthan T et al. Laboratory contamination in airway microbiome studies. BMC microbiology. 2019. PMID 31412780. DOI: 10.1186/s12866-019-1560-1. Access: abstract_metadata. Study: PMID31412780.

33. Zozaya-Valdés E, Wong SQ, Raleigh J et al. Detection of cell-free microbial DNA using a contaminant-controlled analysis framework. Genome biology. 2021. PMID 34162397. DOI: 10.1186/s13059-021-02401-3. Access: abstract_metadata. Study: PMID34162397.

34. Erb-Downward JR, Falkowski NR, D'Souza JC et al. Critical Relevance of Stochastic Effects on Low-Bacterial-Biomass 16S rRNA Gene Analysis. mBio. 2020. PMID 32518181. DOI: 10.1128/mBio.00258-20. Access: abstract_metadata. Study: PMID32518181.

35. Porter AF, Cobbin J, Li CX et al. Metagenomic Identification of Viral Sequences in Laboratory Reagents. Viruses. 2021. PMID 34834931. DOI: 10.3390/v13112122. Access: abstract_metadata. Study: PMID34834931.

36. Rumbavicius I, Rounge TB, Rognes T et al. HoCoRT: host contamination removal tool. BMC bioinformatics. 2023. PMID 37784008. DOI: 10.1186/s12859-023-05492-w. Access: abstract_metadata. Study: PMID37784008.

37. Dohlman AB, Arguijo Mendoza D, Ding S et al. The cancer microbiome atlas: a pan-cancer comparative analysis to distinguish tissue-resident microbiota from contaminants. Cell host & microbe. 2021. PMID 33382980. DOI: 10.1016/j.chom.2020.12.001. Access: abstract_metadata. Study: PMID33382980.

38. Haque MM, Bose T, Dutta A et al. CS-SCORE: Rapid identification and removal of human genome contaminants from metagenomic datasets. Genomics. 2015. PMID 25944184. DOI: 10.1016/j.ygeno.2015.04.005. Access: abstract_metadata. Study: PMID25944184.

39. Liu Y, Elworth RAL, Jochum MD et al. De novo identification of microbial contaminants in low microbial biomass microbiomes with Squeegee. Nature communications. 2022. PMID 36357382. DOI: 10.1038/s41467-022-34409-z. Access: abstract_metadata. Study: PMID36357382.

40. Flores Bueso Y, Walker SP, Hogan G et al. Protoblock - A biological standard for formalin fixed samples. Microbiome. 2020. PMID 32828122. DOI: 10.1186/s40168-020-00901-1. Access: abstract_metadata. Study: PMID32828122.

41. Davis NM, Proctor DM, Holmes SP et al. Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome. 2018. PMID 30558668. DOI: 10.1186/s40168-018-0605-2. Access: abstract_metadata. Study: PMID30558668.

42. Karstens L, Asquith M, Davin S et al. Controlling for Contaminants in Low-Biomass 16S rRNA Gene Sequencing Experiments. mSystems. 2019. PMID 31164452. DOI: 10.1128/mSystems.00290-19. Access: full_text_partial. Study: PMID31164452.

43. Lazarevic V, Gaïa N, Girard M et al. Decontamination of 16S rRNA gene amplicon sequence datasets based on bacterial load assessment by qPCR. BMC microbiology. 2016. PMID 27107811. DOI: 10.1186/s12866-016-0689-4. Access: abstract_metadata. Study: PMID27107811.

44. Merker Breyer G, De Carli S, Rocha Jacques Da Silva ME et al. Alternative amplicon-PCR protocol for maximizing bacterial and fungal sequencing in low-biomass samples. Analytical biochemistry. 2024. PMID 38145697. DOI: 10.1016/j.ab.2023.115449. Access: abstract_metadata. Study: PMID38145697.

45. Rinke C, Low S, Woodcroft BJ et al. Validation of picogram- and femtogram-input DNA libraries for microscale metagenomics. PeerJ. 2016. PMID 27688978. DOI: 10.7717/peerj.2486. Access: abstract_metadata. Study: PMID27688978.

46. Minich JJ, Sanders JG, Amir A et al. Quantifying and Understanding Well-to-Well Contamination in Microbiome Research. mSystems. 2019. PMID 31239396. DOI: 10.1128/mSystems.00186-19. Access: abstract_metadata. Study: PMID31239396.

47. Dyrhovden R, Rippin M, Øvrebø KK et al. Managing Contamination and Diverse Bacterial Loads in 16S rRNA Deep Sequencing of Clinical Samples: Implications of the Law of Small Numbers. mBio. 2021. PMID 34101489. DOI: 10.1128/mBio.00598-21. Access: abstract_metadata. Study: PMID34101489.

48. Caruso V, Song X, Asquith M et al. Performance of Microbiome Sequence Inference Methods in Environments with Varying Biomass. mSystems. 2019. PMID 30801029. DOI: 10.1128/mSystems.00163-18. Access: abstract_metadata. Study: PMID30801029.

49. Stinson LF, Keelan JA, Payne MS et al. Identification and removal of contaminating microbial DNA from PCR reagents: impact on low-biomass microbiome analyses. Letters in applied microbiology. 2019. PMID 30383890. DOI: 10.1111/lam.13091. Access: abstract_metadata. Study: PMID30383890.

50. Stebner A, Ensser A, Geißdörfer W et al. Molecular diagnosis of polymicrobial brain abscesses with 16S-rDNA-based next-generation sequencing. Clinical microbiology and infection : the official publication of the European Society of Clinical Microbiology and Infectious Diseases. 2021. PMID 32244052. DOI: 10.1016/j.cmi.2020.03.028. Access: abstract_metadata. Study: PMID32244052.

51. Son H, Moon J, Ha EJ et al. Identification of bacterial pathogens in brain abscesses by metagenomic approach using nanopore 16S amplicon sequencing. Diagnostic microbiology and infectious disease. 2023. PMID 37741170. DOI: 10.1016/j.diagmicrobio.2023.116041. Access: abstract_metadata. Study: PMID37741170.

52. Al Masalma M, Armougom F, Scheld WM et al. The expansion of the microbiological spectrum of brain abscesses with use of multiple 16S ribosomal DNA sequencing. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2009. PMID 19335164. DOI: 10.1086/597578. Access: abstract_metadata. Study: PMID19335164.

53. Mbongo W, Laga AC, Solomon IH et al. Detection of Gram-positive and Gram-negative bacteria in brain abscesses by 16S rRNA in situ hybridization. Journal of neuropathology and experimental neurology. 2025. PMID 39509244. DOI: 10.1093/jnen/nlae118. Access: abstract_metadata. Study: PMID39509244.

54. Dreses-Werringloer U, Bhuiyan M, Zhao Y et al. Initial characterization of Chlamydophila (Chlamydia) pneumoniae cultured from the late-onset Alzheimer brain. International journal of medical microbiology : IJMM. 2009. PMID 18829386. DOI: 10.1016/j.ijmm.2008.07.002. Access: abstract_metadata. Study: PMID18829386.

55. Hammond CJ, Hallock LR, Howanski RJ et al. Immunohistological detection of Chlamydia pneumoniae in the Alzheimer's disease brain. BMC neuroscience. 2010. PMID 20863379. DOI: 10.1186/1471-2202-11-121. Access: abstract_metadata. Study: PMID20863379.

56. Dominy SS, Lynch C, Ermini F et al. Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Science advances. 2019. PMID 30746447. DOI: 10.1126/sciadv.aau3333. Access: abstract_metadata. Study: PMID30746447.

57. Lei S, Li J, Yu J et al. Porphyromonas gingivalis bacteremia increases the permeability of the blood-brain barrier via the Mfsd2a/Caveolin-1 mediated transcytosis pathway. International journal of oral science. 2023. PMID 36631446. DOI: 10.1038/s41368-022-00215-y. Access: abstract_metadata. Study: PMID36631446.

58. Gschwind R, Fournier T, Kennedy S et al. Evidence for contamination as the origin for bacteria found in human placenta rather than a microbiota. PloS one. 2020. PMID 32776951. DOI: 10.1371/journal.pone.0237232. Access: abstract_metadata. Study: PMID32776951.

59. Theis KR, Romero R, Winters AD et al. Lack of Evidence for Microbiota in the Placental and Fetal Tissues of Rhesus Macaques. mSphere. 2020. PMID 32376701. DOI: 10.1128/mSphere.00210-20. Access: abstract_metadata. Study: PMID32376701.

Source-specific evidence cards

[1] Brain microbial populations in HIV/AIDS: α-proteobacteria predominate independent of host immune status.

Shows: Surgical epilepsy and autopsy brain show bacterial RNA, glial antigen/ISH and seven-week RNA recovery after implantation in Rag1-null mice, with PCR water and tool controls.

Does not show: No human culture or resident-community disease phenotype; heated ODC13 and unheated ODC14 are different donors, confounding the viability contrast. SURG is epilepsy tissue, not healthy population sampling.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Results:paragraph15, 1:Results:paragraph17, 1:Results:paragraph18, 1:Results:paragraph19, 1:Results:paragraph23, 1:Results:paragraph24, 3:Materials and Methods:paragraph4, 3:Materials and Methods:paragraph5, 3:Materials and Methods:paragraph10. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3.

[2] Much ado about nothing? Off-target amplification can lead to false-positive bacterial brain microbiome detection in healthy and Parkinson's disease individuals.

Shows: 47 postmortem and three surgical samples were analyzed with extraction/PCR blanks, mock dilutions, qPCR and host alignment; disease-differentiating taxa were host off-targets.

Does not show: Null is bounded to this assay/cohort. Positive mock missed Listeria and under-recovered Bacillus without mechanical lysis; residual human bacterial load was not literally zero.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Methods:paragraph0, 1:Methods:paragraph1, 1:Methods:paragraph15, 1:Methods:paragraph23, 1:Methods:paragraph26, 1:Methods:paragraph28, 2:Results:paragraph2, 2:Results:paragraph15, 2:Results:paragraph16, 2:Results:paragraph18, 2:Results:paragraph20, 2:Results:paragraph24, 2:Results:paragraph27, 2:Results:paragraph31. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[3] 16S rRNA Next Generation Sequencing Analysis Shows Bacteria in Alzheimer's Post-Mortem Brain.

Shows: AD temporal tissue yields more normalized 16S reads than controls; original Discussion acknowledges human 18S coamplification and possible secondary infection.

Does not show: Normalized library read counts do not quantify absolute organisms or viability; age/postmortem matching cannot alone exclude reagent or host artifacts.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Materials and Methods:paragraph0, 1:Materials and Methods:paragraph5, 1:Materials and Methods:paragraph6, 1:Materials and Methods:paragraph7, 1:Materials and Methods:paragraph11, 3:Discussion:paragraph0, 3:Discussion:paragraph2, 3:Discussion:paragraph3, 3:Discussion:paragraph9. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[4] High resolution 16S rRNA gene Next Generation Sequencing study of brain areas associated with Alzheimer's and Parkinson's disease.

Shows: High-resolution 16S profiling reports bacterial differences across AD/PD-associated brain regions and peripheral comparator tissues.

Does not show: Taxonomic detection cannot establish living parenchymal residence, temporal direction or causal disease contribution; abstract-only controls.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[5] Microspatial Heterogeneities and the Absence of Postmortem Contamination in Alzheimer's Disease Brain Microbiota: An Alzheimer's Pathobiome Initiative (AlzPI) Study.

Shows: 86 samples from 31 specimens, 14 AD/17 controls, with four extraction blanks and edge/core sampling; no edge enrichment and some disease-associated taxa after decontamination.

Does not show: No edge enrichment is not statistical equivalence or absence of all contamination; both edge and core may share reagent/intravascular sources. No culture or prospective disease ordering.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:2. Materials and Methods:paragraph0, 1:2. Materials and Methods:paragraph4, 1:2. Materials and Methods:paragraph5, 1:2. Materials and Methods:paragraph6, 2:3. Results and Discussion:paragraph0, 2:3. Results and Discussion:paragraph3, 2:3. Results and Discussion:paragraph4, 2:3. Results and Discussion:paragraph6, 2:3. Results and Discussion:paragraph8. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[6] Evidence supportive of a bacterial component in the etiology for Alzheimer's disease and for a temporal-spatial development of a pathogenic microbiome in the brain.

Shows: 130 samples from 32 subjects and four extraction blanks support full-length 16S patterns; LDA extracts co-occurrence classes.

Does not show: LDA classes are not organism abundance or measured longitudinal time. Ordering assumes health precedes disease; inferred stages do not establish microbial initiation.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Materials and methods:paragraph0, 1:Materials and methods:paragraph25, 1:Materials and methods:paragraph44, 1:Materials and methods:paragraph49. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3.

[7] Regional Differences in Microbial Infiltration of Brain Tissue from Alzheimer's Disease Patients and Control Individuals.

Shows: Bacterial products and microbial profiles vary by region in AD/control tissue, including widespread LPS immunoreactivity.

Does not show: LPS indicates material, not viability, and region differences permit entry/clearance or handling explanations; original body not examined.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[8] Brain Microbiota in Huntington's Disease Patients.

Shows: Seven Huntington disease brains show fungal/bacterial structures and sequence signals across regions.

Does not show: Same laboratory platforms recur; no contamination-resistant independent living-human community replication or causal intervention.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[9] Different Brain Regions are Infected with Fungi in Alzheimer's Disease.

Shows: Ten AD brains show fungal-antibody immunoreactivity in different regions alongside sequence-based evidence.

Does not show: Antibody structures and postmortem fungal DNA do not identify a reproducible viable community in living healthy brains.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[10] Identification of Fungal Species in Brain Tissue from Alzheimer's Disease by Next-Generation Sequencing.

Shows: Fungal ITS sequencing in nine AD subjects identifies diverse species and regional differences.

Does not show: Amplified low-biomass taxonomic signals require matched blanks and viability; relative diversity does not establish causality.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[11] Polymicrobial Infections In Brain Tissue From Alzheimer's Disease Patients.

Shows: AD tissue shows polymicrobial antibody/PCR signals; Borrelia/chlamydial antibodies cross-react and targeted Borrelia/HSV PCR is negative in the tested series.

Does not show: Cross-reactive immunoreactivity cannot identify those pathogens; positive broad taxonomies and negative specific PCR test different targets.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[12] Parkinson's Disease: A Comprehensive Analysis of Fungi and Bacteria in Brain Tissue.

Shows: Six PD subjects show fungal and bacterial signals by immunohistology and sequencing.

Does not show: Small postmortem series from recurring platform, without prospective viability or removal/rescue; disease-associated entry remains possible.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[13] Corpora Amylacea of Brain Tissue from Neurodegenerative Diseases Are Stained with Specific Antifungal Antibodies.

Shows: Corpora amylacea from several neurodegenerative diseases stain with antifungal antibodies.

Does not show: Storage/clearance bodies and antigens do not demonstrate live microbes or disease-specific causal residence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[14] Evidence for fungal infection in cerebrospinal fluid and brain tissue from patients with amyotrophic lateral sclerosis.

Shows: ALS CSF and brain contain fungal protein/DNA signals in reported assays.

Does not show: CSF and postmortem products are not living parenchymal community evidence; no phenotype-specific eradication.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[15] Direct visualization of fungal infection in brains from patients with Alzheimer's disease.

Shows: Intracellular fungal-like staining occurs in AD neurons, compared with experimentally Candida-infected animals.

Does not show: Infected-animal morphology supports assay plausibility but cannot identify viable human species or distinguish nonspecific staining.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[16] Fungal Enolase, β-Tubulin, and Chitin Are Detected in Brain Tissue from Alzheimer's Disease Patients.

Shows: Fungal enolase, tubulin and chitin antibody signals appear in AD tissue, with some control positives and chitin-negative corpora amylacea.

Does not show: Multiple antigens are not culture or a strain-resolved community; heterogeneous negative/positive structures limit specificity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[17] Infection of Fungi and Bacteria in Brain Tissue From Elderly Persons and Patients With Alzheimer's Disease.

Shows: Fungal/bacterial profiling reports AD-control differences; original Methods describe rapid autopsy, sterile sampling and multiple PCR negatives.

Does not show: Negatives are documented, contrary to a blanket absent-controls claim. Methods examined do not establish independent sequenced bacterial extraction blanks, quantitative load, host screening or viability.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Materials and Methods:paragraph1, 1:Materials and Methods:paragraph4, 1:Materials and Methods:paragraph5, 1:Materials and Methods:paragraph8, 1:Materials and Methods:paragraph9. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[18] Human and Microbial Proteins From Corpora Amylacea of Alzheimer's Disease.

Shows: Two AD corpora-amylacea preparations contain reported microbial and host proteins, with DNA assays supporting fungal signals.

Does not show: Protein attribution in small selected preparations does not establish living microbial residence or its necessity for disease.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[19] Cerebrospinal Fluid from Alzheimer's Disease Patients Contains Fungal Proteins and DNA.

Shows: AD CSF slot-blot/PCR assays report fungal material.

Does not show: Fluid material need not reflect a viable parenchymal community; assay contamination and specificity need original control scrutiny.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[20] Fungal infection in patients with Alzheimer's disease.

Shows: AD specimens show fungal protein/PCR signals interpreted as infection.

Does not show: Molecules can persist after clearance or enter from elsewhere; infection is distinct from a stable resident community.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[21] Gram-negative bacterial molecules associate with Alzheimer disease pathology.

Shows: 24 AD and 18 control brains contain LPS/E. coli K99 immunoreactivity, with stronger signals associated with AD pathology.

Does not show: Control positivity and pathology correlation do not establish viable bacteria, causal direction or resident-community necessity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3.

[22] Bacterial Dissemination to the Brain in Sepsis.

Shows: Sepsis mice have viable brain bacteria around day five but not day 14; human postmortem sepsis brain signals correlate with the systemic source.

Does not show: Demonstrates transient disease-associated dissemination, not homeostatic residency or sporadic neurodegenerative causality.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3.

[23] A brain microbiome in salmonids at homeostasis.

Shows: Perfused trout brains show RNA/DNA, localized bacteria and 54 cultured isolates; buffer/CSF negatives, environment controls, antibiotics and tagged strain colonization corroborate fish residence.

Does not show: Not human evidence. Tagged-strain persistence declines by day 14, brain isolates come from two fish for genome comparisons, and adult salmon LPS/amyloid association is not a causal degeneration intervention.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:RESULTS:paragraph0, 1:RESULTS:paragraph2, 1:RESULTS:paragraph3, 1:RESULTS:paragraph10, 1:RESULTS:paragraph15, 1:RESULTS:paragraph16, 1:RESULTS:paragraph24. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3.

[24] Cerebrospinal Fluid from Healthy Pregnant Women Does Not Harbor a Detectable Microbial Community.

Shows: 23 healthy pregnant women, 116 CSF/control samples and paired metatranscriptomics show no active viable microbial community after contamination correction.

Does not show: CSF is a distinct compartment and pregnancy a restricted population; cannot prove universal absence of parenchymal organisms.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[25] Distinctive thanatomicrobiome signatures found in the blood and internal organs of humans.

Shows: Four cadavers show tissue-specific postmortem microbial profiles and extraction-related differences.

Does not show: Postmortem presence cannot be back-extrapolated to living residence; tiny comparator does not explain every brain signal.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[26] Effects of Extended Postmortem Interval on Microbial Communities in Organs of the Human Cadaver.

Shows: 40 cadavers sampled over 24–432 hours show organ- and postmortem-interval-associated communities.

Does not show: Long-delay autopsy dynamics do not imply all rapid-autopsy or surgical positives are postmortem invasion.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[27] Investigation of Potential Brain Microbiome in Alzheimer's Disease: Implications of Study Bias.

Shows: Two AD cohorts show processing/postmortem effects greater than the limited AD-control differences.

Does not show: DNA may still be present; the null concerns disease signatures and study bias, not an absolute sterility proof.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[28] Microbial molecule ingress promotes neuroinflammation and brain CCR5 expression in persons with HIV-associated neurocognitive disorders.

Shows: 22 HIV/11 control human and 11 SIV/4 control primate specimens show microbial molecules; peptidoglycan alters CCR5/HIV and neuronal mitochondrial readouts.

Does not show: Translocated microbial molecules can act without resident viable bacteria; HIV-linked immune/barrier pathology is not sporadic neurodegeneration. Online 2022, issue 2023.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[29] Reagent and laboratory contamination can critically impact sequence-based microbiome analyses.

Shows: Serial pure-culture dilutions across three facilities and multiple kits produce kit-specific false profiles; cohort-age clustering tracked kit batches.

Does not show: Demonstrates a causal artifact mechanism, not proof that every positive human brain signal is contaminated; tested quantitative floors are protocol-specific.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Results:paragraph0, 1:Results:paragraph4, 1:Results:paragraph8, 1:Results:paragraph9, 1:Results:paragraph10, 1:Results:paragraph11, 1:Results:paragraph13. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[30] Laboratory contamination over time during low-biomass sample analysis.

Shows: 144 negative controls over five years show kit/laboratory/researcher/month variation in contaminant communities.

Does not show: A historical contaminant list cannot classify every future tissue taxon; this is methods evidence, not direct human-brain absence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[31] Separating the signal from the noise in metagenomic cell-free DNA sequencing.

Shows: Low-biomass metagenomic decontamination uses coverage and batch variation and is tested in clinical cfDNA settings.

Does not show: Algorithmic benchmarking in urine/amniotic settings does not measure brain viability or guarantee filtering accuracy there.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[32] Laboratory contamination in airway microbiome studies.

Shows: 23-subject airway study attributes substantial low-biomass profiles to laboratory contamination.

Does not show: Airway-specific fractions do not estimate the brain fraction; matched brain recovery controls remain necessary.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[33] Detection of cell-free microbial DNA using a contaminant-controlled analysis framework.

Shows: A contaminant-controlled workflow detects microbial cfDNA exceeding blanks in plasma samples.

Does not show: Circulating DNA can exist without live resident plasma organisms and may contribute to brain tissue signals.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[34] Critical Relevance of Stochastic Effects on Low-Bacterial-Biomass 16S rRNA Gene Analysis.

Shows: Mock serial dilutions and clinical respiratory samples reveal stochastic profiles below the tested bacterial-load range.

Does not show: The numerical copy threshold is assay-specific and cannot be imposed as a universal brain cutoff.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[35] Metagenomic Identification of Viral Sequences in Laboratory Reagents.

Shows: Seven blank libraries contain 14 viral contaminants, demonstrating reagent-derived RNA/viral reads.

Does not show: RNA detection alone cannot establish viability, but contamination in these reagents is not a universal explanation of all viral tissue findings.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[36] HoCoRT: host contamination removal tool.

Shows: Synthetic/public benchmarks assess host-removal accuracy; long-read performance remains incomplete.

Does not show: A host-removal tool is not direct biological evidence for or against brain residence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[37] The cancer microbiome atlas: a pan-cancer comparative analysis to distinguish tissue-resident microbiota from contaminants.

Shows: Pan-cancer profiles use center/batch and matched tissue information to distinguish candidate tissue signals from contaminants.

Does not show: Cancer/mucosal tissue validation does not establish brain microbiota or living strain identity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[38] CS-SCORE: Rapid identification and removal of human genome contaminants from metagenomic datasets.

Shows: CS-SCORE benchmarks human-read removal with computational efficiency benefits.

Does not show: Speed and genome filtering do not certify low-copy bacterial identification, tissue localization or viability.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[39] De novo identification of microbial contaminants in low microbial biomass microbiomes with Squeegee.

Shows: Squeegee cross-sample contaminant identification is benchmarked against control-based approaches.

Does not show: Data-driven decontamination does not replace independent physical blanks or prove brain community existence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[40] Protoblock - A biological standard for formalin fixed samples.

Shows: Defined mammalian/bacterial FFPE blocks compare recovery with qPCR, 16S and whole-genome methods.

Does not show: Useful assay recovery standard; fixed mock blocks are not viable organisms in human brain.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[41] Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data.

Shows: decontam uses frequency/prevalence with culture dilution and biological cohorts to identify contaminants.

Does not show: Model assumptions can fail with cross-contamination or shared true/background taxa; filtered signal is not automatically true residence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[42] Controlling for Contaminants in Low-Biomass 16S rRNA Gene Sequencing Experiments.

Shows: Known mock dilutions quantify contamination and false deletion; blank-only filtering removes 20.1–33.9% of true ASVs and misses many contaminants.

Does not show: No algorithm is perfect across mixtures. Unknown-source SourceTracker fails substantially, while known-source success cannot be assumed for brain.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:RESULTS:paragraph12, 1:RESULTS:paragraph17, 1:RESULTS:paragraph18, 1:RESULTS:paragraph20, 1:RESULTS:paragraph22, 1:RESULTS:paragraph25. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[43] Decontamination of 16S rRNA gene amplicon sequence datasets based on bacterial load assessment by qPCR.

Shows: qPCR bacterial-load-aware removal preserves known culture members and reduces tested contaminants.

Does not show: Some sample cross-contaminants escape blank-based correction; recovery must be tested for the actual brain protocol.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[44] Alternative amplicon-PCR protocol for maximizing bacterial and fungal sequencing in low-biomass samples.

Shows: Alternative two-stage 16S/ITS amplification improves low-biomass sequencing in mocks and clinical samples.

Does not show: Amplification yield or consistency cannot demonstrate microbial viability or absence of stochastic errors below recovery limits.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[45] Validation of picogram- and femtogram-input DNA libraries for microscale metagenomics.

Shows: 54-species mock and environmental microsamples validate very-low-input shotgun libraries with explicit complexity/contamination limits.

Does not show: Real microspatial variation is possible; low-input validation in known communities is not automatic brain validity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[46] Quantifying and Understanding Well-to-Well Contamination in Microbiome Research.

Shows: Unique bacterial assignments across two laboratories reveal well-to-well contamination, strongest at low biomass.

Does not show: Cross-contamination can introduce real tissue taxa into blanks; blanket deletion of blank taxa can therefore erase true organisms.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[47] Managing Contamination and Diverse Bacterial Loads in 16S rRNA Deep Sequencing of Clinical Samples: Implications of the Law of Small Numbers.

Shows: Mock and bile analyses with repeated controls show stochastic amplification and incomplete single-blank coverage.

Does not show: Clinical bile is not brain; quantitative per-sample decision rules require protocol-specific recovery rather than taxon blacklists.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[48] Performance of Microbiome Sequence Inference Methods in Environments with Varying Biomass.

Shows: Six inference approaches across biomass dilutions compare false sequence/OTU calls; ASV methods improve precision.

Does not show: Sequence inference does not by itself distinguish true low-load tissue biology from contamination.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[49] Identification and removal of contaminating microbial DNA from PCR reagents: impact on low-biomass microbiome analyses.

Shows: DNase treatment of PCR master mixes reduces measured reagent contamination about 99% in the tested protocol.

Does not show: Kit extraction is not always the largest contaminant source; treatment must preserve assay sensitivity. Online 2018, issue 2019.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[50] Molecular diagnosis of polymicrobial brain abscesses with 16S-rDNA-based next-generation sequencing.

Shows: 79 samples from 54 patients include 51 intracranial abscesses, with sequencing detecting broader polymicrobial pathogens.

Does not show: Diagnostic infection is a positive assay control, not evidence of a normal resident human brain community.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[51] Identification of bacterial pathogens in brain abscesses by metagenomic approach using nanopore 16S amplicon sequencing.

Shows: Nanopore 16S of 27 abscess pus samples detects more anaerobic/polymicrobial findings than culture.

Does not show: Abscess load, inflammation and collection differ from healthy low-biomass brain; detection is not homeostasis.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[52] The expansion of the microbiological spectrum of brain abscesses with use of multiple 16S ribosomal DNA sequencing.

Shows: Prospective 20-patient abscess comparison identifies 72 sequenced strains versus 22 culture strains and expands species spectrum.

Does not show: Clinically localized infection establishes microbes can enter brain, without normal residence or neurodegenerative community causality.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[53] Detection of Gram-positive and Gram-negative bacteria in brain abscesses by 16S rRNA in situ hybridization.

Shows: 63 abscess cases yield 18 ISH positives; reported 35% sensitivity/93% specificity versus the composite comparator.

Does not show: Negative ISH cannot exclude all infection, and abscess diagnostic performance cannot establish healthy-tissue assay sensitivity. Online 2024, issue 2025.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[54] Initial characterization of Chlamydophila (Chlamydia) pneumoniae cultured from the late-onset Alzheimer brain.

Shows: Two AD brain-derived Chlamydia pneumoniae isolates were cultured through HEp-2 passages and characterized genetically.

Does not show: A substantive viability positive in diseased autopsy tissue; does not establish community residence in living brains, healthy controls or AD necessity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[55] Immunohistological detection of Chlamydia pneumoniae in the Alzheimer's disease brain.

Shows: Five AD/five control immunohistological comparisons localize Chlamydia signals near, but not invariably within, amyloid deposits.

Does not show: Proximity and antibody controls do not establish disease ordering, viable community or microbial causation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

[56] Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors.

Shows: Human AD gingipain signals accompany inoculated-mouse bacterial load and small-molecule inhibition effects on neurodegenerative readouts.

Does not show: Forced pathogen exposure and protease inhibition are not removal/rescue of an endogenous human resident community.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[57] Porphyromonas gingivalis bacteremia increases the permeability of the blood-brain barrier via the Mfsd2a/Caveolin-1 mediated transcytosis pathway.

Shows: Repeated intravenous P. gingivalis exposure changes BBB transcytosis; Cav/Mfsd-related interventions affect barrier readouts.

Does not show: Installed bacteremia can permit entry; it cannot establish stable homeostatic human residence or sporadic-disease primacy.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[58] Evidence for contamination as the origin for bacteria found in human placenta rather than a microbiota.

Shows: 38-woman placenta culture/qPCR and selected sequencing distinguish surface organisms from internal signals matching blanks/database artifacts.

Does not show: Sterile-tissue comparator highlights contamination control, but placenta absence does not prove adult brain absence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[59] Lack of Evidence for Microbiota in the Placental and Fetal Tissues of Rhesus Macaques.

Shows: Cesarean fetal brain/placenta signals do not exceed controls; isolated C. acnes also occurs in controls while positive controls work.

Does not show: Restricted fetal macaque setting cannot settle adult human brain ecology; a valuable organ/method comparator only.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

Documentary record

Every grade on this site is reproducible from its record. The files below are the posted edition exactly as assessed.

• ADVERSARIAL_REVIEW.md
• SHA256.txt
• adversarial-findings.json
• codex-correction-provenance.json
• consensus-addendum.md
• consensus-validation.json
• edition-approval-audit.json
• reader-summary.json
• resident-brain-microbiome.review.md
• resident-brain-microbiome.scorecard.json
• validation-report.json
• validation-round2-2026-10-14.json
• validation.txt

Sources: 59 · load-bearing claims: 63 · verified: 63 · unverified: 0 · contradicted: 0 · schema 5.3