Contributions
Lanthanide Methylotrophy Atlas — Summary built a 293,059-genome atlas of lanthanide-dependent methanol oxidation across the KBase Data Lakehouse pangenome 1. It found eggNOG xoxF (K00114) in 3,690 genomes and mxaF (K14028) in 195 genomes, a global xoxF:mxaF ratio of 18.92 : 1 (Clopper-Pearson 95% CI [13.07, 27.69]; one-sided binomial p = 7.6 × 10⁻²² against the pre-registered 10/11 threshold) 1. The project tested this with three phylogenetic validations, and all three had 95% CI lower bounds above the 10 : 1 threshold 1. Naive genome-level pooling over 3,885 MDH-informative genomes gave an xoxF fraction of 0.950 [95% CI 0.942, 0.956] 1. Equal weighting across 271 GTDB families gave 0.960 [95% CI 0.941, 0.976] 1. A Bayesian binomial GLMM (generalized linear mixed model) with phylum and family random intercepts gave 0.993 [95% CI 0.992, 0.994], corresponding to ~143 : 1 [122, 169] 1. After Benjamini–Hochberg false-discovery-rate (FDR) correction, giving adjusted p-values (p_BH), the project found high xoxF rates in Acidobacteriota (28.3%, 95 : 1, p_BH = 1.2 × 10⁻⁷⁹) and Gemmatimonadota (25.4%, 98 : 1, p_BH = 1.5 × 10⁻²⁷) 1. The report noted that annotation-based detection does not establish enzyme activity or methylotrophic physiology 1.
The same project found Bakta-validated lanmodulin in 62 genomes, all within Beijerinckiaceae, Acetobacteraceae or Hyphomicrobiaceae (p = 9.8 × 10⁻⁷ against an 80% threshold) 1. It found xoxF co-occurring with lanmodulin in 49/62 = 79.0% of these genomes, just below the pre-registered 80% co-occurrence threshold, so that hypothesis (H3b) was not formally supported (p = 0.65) 1. It found soil/sediment enriched for xoxF relative to the generic_environmental reference (OR = 1.92, p_BH = 6.1 × 10⁻³⁹) and host-associated samples depleted (OR = 0.058) 1. The rare-earth-element (REE)-impacted signal (n = 37, OR = 3.51, p_BH = 0.082) was descriptively elevated but did not clear the FDR threshold 1. The project found only 4/37 REE-acid-mine-drainage MAGs (metagenome-assembled genomes) carrying any xoxF and 0/37 carrying Bakta-validated lanmodulin, while stress products such as RecN (33/37) were prevalent 1. For PQQ, the obligate XoxF cofactor, it found that 1,288 (59%) of 2,185 xoxF genomes lacking eggNOG pqq annotation had at least 1 Bakta PQQ product, indicating annotation gaps 1. It found 897 xoxF genomes lacking PQQ evidence from either source 1. It also found 505 eggNOG lanM calls that were false positives concentrated in unrelated gut Bacillota 1.
Self-Sufficiency, Anaerobic Toolkit, and Cultivation Bias in Clay-Confined Cultured Bacterial Genomes compared 9 deep-confined clay genomes, 30 shallow-clay genomes and a soil baseline of 150 genomes (137 after quality filtering) 2. It found dissimilatory sulfate-reduction markers in 5/9 deep genomes versus 0.018 expected under the Mitzscherling et al. (2023) rock-attached null (binomial p = 4.0×10⁻¹²) 2. The report read this as matching the Bagnoud porewater paradigm and as a diagnostic of cultivation bias 2. The project found the deep cohort enriched for Wood–Ljungdahl and group 1 [NiFe]-hydrogenase (both p_BH = 0.004 after Benjamini–Hochberg false-discovery-rate correction), but within Bacillota_B neither difference was significant (BH-adjusted p = 1.0) 2. Within Bacillota_B, it found sulfate reduction still associated with deep clay (5/5 versus 4/19, p_BH reported as 0.04 and 0.044) 2. The project withdrew its original iron-reduction narrative because K07811, K17324 and K17323 had been misidentified as iron-reduction markers 2. Its corrected multi-heme cytochrome detector gave rates of 55.6% (n = 9), 40.0% (n = 30) and 40.9% (n = 149) for deep, shallow and baseline cohorts, with all Fisher p ≥ 0.46 2. The project found GapMind amino-acid pathway completeness not elevated in the deep cohort: after CheckM filtering, 6 deep genomes averaged 15.50/18 versus 17.14/18 for 137 baseline genomes (p = 0.009, d = −0.84) 2. The report noted an 18-pathway ceiling and stated that the result does not reject self-sufficiency as a deep-subsurface adaptation 2.
Subsurface Bacillota_B Specialization — What Distinguishes Deep-Clay Lineages from Soil Congeners? compared 10 deep-clay Bacillota_B genomes with 62 soil-baseline Bacillota_B genomes 3. Its Fisher tests across 14,109 eggNOG orthologous groups (OGs) found 547 enriched OGs at q<0.05 3. Manual inspection found COG1977 (molybdopterin cofactor metabolism) in 10/10 anchors versus 11/62 baselines 3. It also found a DsrE/DsrF/DsrH-like OG in 7/10 anchors versus 0/62 baselines 3. The report estimated the H1-relevant total at 80–100 OGs, which it stated is not a finalized count 3. The project rejected the prediction that deep-clay genomes would be smaller, in the opposite direction 3. Anchor genomes averaged 4,110,038 bp versus 3,046,124 bp (Cohen’s d = +1.39, p = 0.025) 3. CheckM completeness was 94.7% versus 94.3% (p = 0.93), so the report concluded a genome-quality artifact did not explain the difference 3. The project’s corrected multi-heme cytochrome comparison found no significant cohort differences (e.g. deep versus shallow OR 1.88, p = 0.46) 3. It found that multi-heme cytochrome PFAM protein-family domains were sparse within Bacillota_B (~6 clusters), so CXXCH motif counting carried most of the corrected signal 3. The report cautioned that marginal effects with anchor counts of 3–5 should be treated descriptively, and that spanning four orders only partly mitigates genus-level phylogenetic confounding 3.