Biology subjects

Whalley, J. P.

Publications and source records attributed to Whalley, J. P..

4 recordsLinked to original sources

bioRxiv2026

Glitch genes: embedding geometry predicts functional fragility in single-cell foundation models

Single-cell foundation models (scFMs) produce gene embeddings that are increasingly interpreted biologically, but it is unclear whether different models organise gene space consistently, or whether geometric extremes identify important genes. The same four-metric screen was applied to Geneformer, scGPT and scFoundation, which differ in architecture, training objective and expression encoding. The models agreed weakly on individual outlier genes but showed structured class-level convergence: ribosomal enrichment occurred in all three, although weaker and caller-dependent in scFoundation, while strong mitochondrial enrichment was confined to Geneformer and scGPT. Gene-level agreement was itself uneven, with Geneformer and scGPT overlapping above chance and scGPT and scFoundation not, and caller stability differed between models. Most outliers were not extreme in ESM-2 protein-sequence space, indicating principally model-specific rather than protein-sequence geometry. These patterns did not translate into the tested notions of gene importance: deleting the highest-anomaly Geneformer genes did not impair cell-type annotation beyond matched controls, and covariate-adjusted outlier status was not associated with ClinVar membership in any model. Gene-embedding geometry therefore characterises model-specific structure, but does not provide standalone evidence of downstream leverage or disease relevance.

bioinformatics↗

bioRxiv2025

Aged brain multi-omic integration captures immunometabolic and sex variation

Resolving the molecular basis for heterogeneous ageing in the human brain requires integrating its diverse molecular layers. Here we applied a tensor-decomposition framework to jointly analyse single-nucleus RNA-seq, DNA methylation, H3K9ac histone acetylation and proteomic data from 276 post-mortem human cortices. The analysis uncovered two dominant, multi-omic themes. First, we identified an immunometabolic axis characterised by microglial activation, suppression of PI3K-Akt-mTOR signalling and epigenetic signatures that suggest repurposing of developmental transcription factors such as BARHL1. Second, we decomposed the effect of sex into a multi-omic program linked to cellular plasticity: although driven by canonical sex-chromosome genes, this axis was epigenetically enriched for pathways regulating stem-cell pluripotency and exhibited a female-biased shift toward precursor-like states in the oligodendrocyte lineage. Together, these findings deliver a high-resolution atlas of glial ageing and demonstrate that broad risk factors manifest as complex, cell-specific vulnerabilities; an insight critical for the development of targeted therapeutic strategies.

neuroscience↗

bioRxiv2021

Using de novo assembly to identify structural variation of complex immune system gene regions

Driven by the necessity to survive environmental pathogens, the human immune system has evolved exceptional diversity and plasticity, to which several factors contribute including inheritable structural polymorphism of the underlying genes. Characterizing this variation is challenging due to the complexity of these loci, which contain extensive regions of paralogy, segmental duplication and high copy-number repeats, but recent progress in long-read sequencing and optical mapping techniques suggests this problem may now be tractable. Here we assess this by using long-read sequencing platforms from PacBio and Oxford Nanopore, supplemented with short-read sequencing and Bionano optical mapping, to sequence DNA extracted from CD14+ monocytes and peripheral blood mononuclear cells from a single European individual identified as HV31. We use this data to build a de novo assembly of eight genomic regions encoding four key components of the immune system, namely the human leukocyte antigen, immunoglobulins, T cell receptors, and killer-cell immunoglobulin-like receptors. Validation of our assembly using k-mer based and alignment approaches suggests that it has high accuracy, with estimated base-level error rates below 1 in 10 kb, although we identify a small number of remaining structural errors. We use the assembly to identify heterozygous and homozygous structural variation in comparison to GRCh38. Despite analyzing only a single individual, we find multiple large structural variants affecting core genes at all three immunoglobulin regions and at two of the three T cell receptor regions. Several of these variants are not accurately callable using current algorithms, implying that further methodological improvements are needed. Our results demonstrate that assessing haplotype variation in these regions is possible given sufficiently accurate long-read and associated data; application of these methods to larger samples would provide a broader catalogue of germline structural variation at these loci, an important step toward making these regions accessible to large-scale genetic association studies.

bioinformatics↗

bioRxiv2017

Framework For Quality Assessment Of Whole Genome, Cancer Sequences

Working with cancer whole genomes sequenced over a period of many years in different sequencing centres requires a validated framework to compare the quality of these sequences. The Pan-Cancer Analysis of Whole Genomes (PCAWG) of the International Cancer Genome Consortium (ICGC), a project a cohort of over 2800 donors provided us with the challenge of assessing the quality of the genome sequences. A non-redundant set of five quality control (QC) measurements were assembled and used to establish a star rating system. These QC measures reflect known differences in sequencing protocol and provide a guide to downstream analyses of these whole genome sequences. The resulting QC measures also allowed for exclusion samples of poor quality, providing researchers within PCAWG, and when the data is released for other researchers, a good idea of the sequencing quality. For a researcher wishing to apply the QC measures for their data we provide a Docker Container of the software used to calculate them. We believe that this is an effective framework of quality measures for whole genome, cancer sequences, which will be a useful addition to analytical pipelines, as it has to the PCAWG project.

genomics↗