Biology subjects

Chen, L. G.

Publications and source records attributed to Chen, L. G..

3 recordsLinked to original sources

bioRxiv2026

LRCQ: phenome-wide heritability enrichment of genomic loci from GWAS summary statistics

Some genetic variants are associated with many traits, but a variant's GWAS statistics are also inflated by the effects of variants in LD with it and by confounding shared by all variants in a study. Existing measures of per-variant pleiotropy, such as counts of significant associations or multi-trait association tests, do not separate these sources. Stratified LD score regression separates them for one trait at a time and only through predefined annotations. We present LRCQ (LD score regression on {chi}2 across many GWAS), which estimates for each locus its phenome-wide heritability enrichment: the variance of its causal effects across a set of traits, relative to an average locus. LRCQ regresses squared Z-scores from 30 or more GWAS jointly on squared LD, with per-trait LDSC intercepts and heritabilities fitted in earlier stages. Its pooled least-squares solution has a closed form whose cost does not grow with the number of traits. Its standard errors account for sample overlap and genetic correlation between GWAS. Squared Z-scores identify enrichment at the resolution of LD clumps, not single SNPs, so LRCQ reports clump-level estimates ("omni-loci"). These estimates can be aggregated over any functional annotation or gene set chosen after the analysis. In simulations with real UK Biobank LD and 30 to 1,000 traits, LRCQ was unbiased for its clump-level estimand. When GWAS share samples its test statistics are strongly right-skewed: normal-reference p-values were up to 10-fold anti-conservative at P < 0.001, and skewness-corrected p-values from the exact third cumulant restored type-I error and FDR control. Model-based standard errors were calibrated once the enrichment estimates plugged into them kept the polygenic background. It ranked enriched loci better than mean {chi}2, LD-normalised {chi}2, multi-trait omnibus tests and significant-association counts in 13 of 16 scenarios. Its main vulnerability was a misspecified LDSC intercept. Applied to 444 Pan-UK Biobank European-ancestry GWAS, LRCQ found strongly non-uniform phenome-wide enrichment: the SD of true clump-mean enrichment was 8.3, against a genome mean of 1. It identified 70 omni-loci (99 LD clumps) at FDR 5%, including CRP, CETP, APOB, AKTIP, SMOX, the SHBG region, CASZ1 and HNF1A. These ranked only moderately with mean {chi}2 and poorly with counts of significant associations. Gene-level phenome-wide enrichment rose with loss-of-function constraint. Phenome-wide heritability was 3-5-fold enriched in coding, conserved, 5' UTR and promoter sequence, and depleted in repressed chromatin. LRCQ turns hundreds of biobank GWAS into an LD-corrected map of where phenome-wide genetic variance lies, and complements its companion method LRCP, which estimates genetic-effect correlations between loci.

genetics↗

bioRxiv2026

LRCP: identifying functionally related SNPs and genes from phenome-wide genetic-effect correlations using GWAS summary statistics

Genetic correlation describes which traits share genetic causes. The transposed question, which variants act alike across the phenome, has no comparable estimator, although it points directly to shared genes, pathways and regulators. Correlations of GWAS Z scores between variants across many traits are easy to compute, but they are confounded by linkage disequilibrium (LD) and, for any pair of variants, by sample overlap and genetic correlation among the GWAS. We show that in 452 Pan-UK Biobank GWAS the effective number of independent traits is only 58, and that naive tests reject 51-70% of null pairs of distal variants. We introduce LRCP (likelihood and regression on cross-products of Z scores), which estimates the phenome-wide genetic-effect correlation{rho} between SNPs by regressing Z-score cross-products on LD cross components, using per-SNP heritability enrichment from the companion method LRCQ and LDSC-derived plug-ins. For variants in different LD blocks the estimator has a closed form, and its standard errors account exactly for the dependence among GWAS. In simulations LRCP is unbiased (absolute bias at most 0.03), its 95% intervals cover 93-98%, and its type-I error is 5.0-5.7%, whereas naive correlations are attenuated by 41-64% and reject up to 51% of null pairs. In Pan-UK Biobank, LRCP recovers the expected signs among PCSK9, LDLR and APOB variants and estimates{rho} = 0.90 between the TP53 polyadenylation variant rs78378222 and MDM2 rs3730556, whose Z scores correlate at r = 0.67 across traits (conditional bootstrap P = 0.015), although this correlation lies along a shared body-size and red-cell trait axis. Single variant pairs are near the resolution of current biobanks, with precision limited by the number of genetically distinct traits. Genome-wide, among 7,793 enriched LD clumps (30.2 million distal pairs), a conditional test finds 17 pairs at FDR 5%, where naive correlation flags 23 million. Simulating the whole screen under the null shows that this test is conservative and that the 113 strongest links (P < 10-5) lie far beyond the null, which yields 0.025 such pairs per screen. Linked genes share Reactome pathways 4.9 times more often than random pairs of enriched loci. The top pairs form 19 modules, the largest of which match known biology: myeloid differentiation (IRF8, CEBPE, IKZF1), LDL and HDL metabolism, iron homeostasis (TMPRSS6, TFRC), the red-cell membrane (SPTA1, ANK1) and type-2 immunity (TSLP, IL13). LRCP turns phenome-wide summary statistics into a map of functionally related variants and genes.

genetics↗

bioRxiv2023

Challenges in the discovery of tumor-specific alternative splicing-derived cell-surface antigens in glioma

BackgroundDespite advancements in cancer immunotherapy, solid tumors remain formidable challenges. In glioma, profound inter-and intra-tumoral heterogeneity of antigen landscape hampers therapeutic development. Therefore, it is critical to consider alternative sources to expand the repertoire of targetable (neo-)antigens and improve therapeutic outcomes. Accumulating evidence suggests that tumor-specific alternative splicing (AS) could be an untapped reservoir of neoantigens. ResultsIn this study, we investigated tumor-specific AS events in glioma, focusing on those predicted to generate major histocompatibility complex (MHC)-presentation-independent, cell-surface neoantigens that could be targeted by antibodies and chimeric antigen receptor (CAR)-T cells. We systematically analyzed bulk RNA-sequencing datasets comparing 429 tumor samples (from The Cancer Genome Atlas [TCGA]) and 9,166 normal tissue samples (from the Genotype-Tissue Expression project [GTEx]), and identified 13 AS events in 7 genes predicted to be expressed in more than 10% of the patients, including PTPRZ1 and BCAN, which were corroborated by an external RNA-sequencing dataset. Subsequently, we validated our predictions and elucidated the complexity of the isoforms using full-length transcript amplicon sequencing on patient-derived glioblastoma cells. However, analyses of the RNA-sequencing datasets of spatially mapped and longitudinally collected clinical tumor samples unveiled remarkable spatiotemporal heterogeneity of the candidate AS events. Furthermore, proteomics analysis did not reveal any peptide spectra matching the putative neoantigens. ConclusionsOur investigation illustrated the diverse characteristics of the tumor-specific AS events and the challenges of antigen exploration due to their notable spatiotemporal heterogeneity and elusive nature at the protein levels. Redirecting future efforts toward intracellular, MHC-presented antigens could offer a more viable avenue.

cancer biology↗