What happened
Improvements in computational protein structure prediction have enabled searches for remote homologs of proteins whose molecular function remain unknown. However, the reliability of functional annotations from such searches has not been systematically quantified. In this study, we assembled a time-split benchmark from Pfam families annotated as domains of unknown function (DUFs) in Pfam 28.0 and were subsequently assigned a function in the current release (Pfam 38.0). These retrospective-DUFs provide ground truth for assessing functional annotation from remote homology searches. In our benchmark analysis, we paired retrospective-DUFs with a difficulty-matched arm from known domains to distinguish query difficulty from the method's performance. Across four model proteomes (yeast, C. elegans, Drosophila, and mice), Foldseek searches against AlphaFold/Swiss-Prot, PDB100, and CATH50 recovered the later-assigned function for 15.9% of retrospective DUF queries, compared with 30.1% of matched known-domain queries, after masking for self-family and self-clan level hits to remove circularity. At a fixed confidence cut-off (qTM [≥] 0.5), 55.4% of informative calls on retrospective DUFs were incorrect, establishing an error model for prospective use. Furthermore, in our comparison of tools for homology searches, Foldseek outperformed MMseq2-based sequence search but was not statistically separable from the ESM-2 protein language model embedding baseline. Applying the calibrated structural homology search pipeline to 296 currently unannotated DUF queries in Pfam 38.0 yielded 50 confident, specific functional assignments. We provide the benchmark and ranked candidates as a resource to facilitate functional studies. Update (2026-09-16): Uncertainty-Aware Model Selection with a Calibrated Probability-Generating-Function-Based Bayesian Information Criterion . Selecting stochastic gene-expression models from single-cell counts requires balancing goodness of fit against unnecessary mechanistic complexity. The probability-generating-function-based Bayesian information criterion (PGF-BIC) combines covariance-weighted fitting in generating-function space with a complexity penalty, allowing candidate models to be compared without reconstructing their full count distributions. However, its conventional zero-threshold rule does not account for sampling uncertainty in the fitted score difference and may therefore favor overly complex models in finite samples. To address this limitation, we develop an uncertainty-aware PGF-BIC rule that selects the more complex model only when its score advantage exceeds a data-driven threshold. We use Cantelli's one-sided inequality to motivate a selection margin expressed in terms of a standard deviation. To determine this scale, we use influence functions to quantify sensitivity to small perturbations in the data distribution and obtain a first-order description Update (2026-09-19): Scaling Functional Annotation Across Proteomes, Pangenomes and Metagenomes with Sma3s v3 . High-throughput sequencing has generated protein datasets whose scale increasingly exceeds the practical limits of conventional functional annotation workflows. We present Sma3s v3, a scalable reimplementation of the Sma3s three-step annotation strategy, which combines transfer from highly similar homologs, orthology-based inference, and functional enrichment among homologous proteins. Sma3s v3 replaces BLAST-based searches with MMseqs2 and introduces parallel processing, reusable SQLite caches, taxonomic filtering, and traceable outputs that retain the evidence underlying each assignment. We evaluated the method on a Vibrio cholerae pangenome comprising 50,415 gene clusters from 11,295 quality-filtered genomes and on a metagenomic catalogue containing 843,935 proteins. After excluding non-informative assignments, Sma3s v3 annotated 30,662 pangenome clusters (60.8%), comparable to InterProScan (60.2%) and exceeding eggNOG-mapper (41.4%), while providing 5,747 annotations not recovered by either comparator. Gene Ontology comparisons showed broad semantic agreement be Update (2026-09-24): Novel human long noncoding RNA responses to Candida auris and its cell-wall components in peripheral blood mononuclear cells . Candida auris is an emerging multidrug-resistant fungal pathogen. However, little is known about the role of long noncoding RNAs (lncRNAs) in the human immune response to this fungus. We reanalyzed RNA-sequencing data from peripheral blood mononuclear cells obtained from three healthy donors. The cells were exposed to live C. auris, purified mannan, purified {beta} glucan, or RPMI control for 4 h and 24 h. We identified 607 putative novel polyadenylated lncRNA loci that were absent from GENCODE Release 48. Of the 449 novel loci tested for differential expression, 102 responded to at least one stimulus. Novel-lncRNA responses were limited at 4 h but increased markedly at 24 h. At 24 h, 78 novel lncRNAs responded to live C. auris, 34 to mannan, and 11 to {beta}-glucan. Many of these responses were specific to a particular stimulus. Among the 102 responsive lncRNAs, 94 were significantly co-expressed with protein coding genes. The coding-gene partners of the most highly connected lncRNAs were associated with innate immunity, defense