Happy to see our method for predicting trans-eQTLs published at @GenomeBiology. Demonstrated on #eQTL data from @GTExPortal. Fun working with @francosimonetti, @SoedingL, @Raktim7879, @KaphleAnubhav and others. Thread for a quick overview. \1.

RT @jkpritch: Sharing this in case it's useful: I find that students often have a hard time conceptualizing quantitative parameters in geno….

RT @HGEN_UChicago: A successful conference for UChicago Human Genetics at #probgen23 💻🧬 Thanks for having us @cshlmeetings @CSHL.

RT @rikhiag: Excited to share our latest work on membrane endocytosis of molecular condensates formed by liquid liquid phase separation, ou….

RT @fabian_theis: Beautiful perspective @NatureMedicine about the impact of the Human Cell Atlas on medicine by @jennyErood with Sarah @tei….

RT @skathire: Does chronic disease need to equal chronic care? No. 👇in @Nature from @VerveTx . One-time base editing Rx (in vivo liver), .d….

RT @joannahowson: Ever wondered if the mitochondrial DNA affects common disease risk? Well take a look at our work out @NatureGenet We re-c….

RT @CellCellPress: In the latest issue! Population-scale tissue transcriptomics maps long non-coding RNAs to complex disease.@sbmontgom @St….

Project funded by @medicine_e, @MBExC_de, @mpi_bpc. #GWAS data from @hakyim lab. And finally, for all the thoughtful suggestions and critiques which significantly improved the manuscript, a huge Thank You to the unknown reviewer! \10.

There are some major limitations of our method, including (i) choice of the L2 prior, and (ii) inability to use standard confounder correction (we proposed an alternate non-linear, unsupervised, hidden confounder correction method using k-nearest neighbors). \9.

Tejaas is a command line tool available with GPLv3 license at \8.

Overlap of these trans-eQTLs with complex trait GWAS SNPs is tissue-specific. For example, heart atrial appendage and LCL trans-eQTLs overlap with SNPs associated with cardiometabolic disease, breast trans-eQTLs with breast cancer, etc. \7.

The trans-eQTLs are enriched to occur in open chromatin, enhancers, and other regulatory regions (matched with corresponding tissues). They occur near known transcription factors and overlap with reporter assay QTLs. \6.

Predicting trans-eQTLs by Tejaas does not depend on the presence of a cis effect, because cis genes are masked before reverse regression. About 20% of trans-eQTLs were found to act via cis-eQTLs, significantly more than expected by random chance. \5.

Our method predicted 18851 unique trans-eQTLs in GTEx: To our knowledge, this is the first systematic large-scale identification of trans-eQTL associations in the GTEx dataset. Trans-eQTLs are tissue-specific. \4.

Instead of following the assumed direction of cause and effect, we proposed to turn around the direction of regression, using gene expression levels (the effects) as covariates and the SNP (the potential cause) as response. We call it "reverse regression". \3

Identifying trans-eQTLs is challenging because of small effect sizes, tissue specificity, and a severe multiple-testing burden. But recent studies suggest >70% genetic variance in gene expression is contributed by trans-eQTLs, and is crucial for understanding disease pathways. \2.

RT @iiscbangalore: The last date for international students to apply for the UG programme 2021-22 at IISc has been extended to May 30, 2021….

RT @yluo86: A game changer - identifying target genes in cis and/or trans for multiple GWAS plausible causal variants in single cells at sc….

RT @drklly: Check out our latest iteration of predicting gene expression and chromatin descriptions from DNA sequence, with a meaningfully….