Eric Kernfeld

@ekernf01.bsky.social

1349 Followers 422 Following

Statistician and computational biologist; uw alum; jhu student. He/him. http://ekernf01.github.io

Photo of Eric with mom, dad, and brother Paul. They are holding flowers and balloons to celebrate after Eric's defense.
Sankey diagram of Eric's job search efforts and opportunities. Source data: Months unemployed [1] Biking Months unemployed [1] Stressing Months unemployed [1] Applying Months unemployed [2] Networking Months unemployed [0] Waiting to joke about "came in a fluffer" Jobs applied to (no referral) [64] No follow up Jobs applied to (with referral) [4] No follow up Jobs applied to (with referral) [3] Interview Jobs applied to (no referral) [1] Interview Info interviews conducted [30] No follow up Info interviews conducted [2] Interview Credible unsolicited inquiries [2] Interview Credible unsolicited inquiries [3] Location mismatch Interview [2] Job offer
Picture of Eric and his bike in front of tons of unstable rock at the mouth of the Paw-Paw Tunnel, part of the C&O Canal Towpath national park
Classic "midwit meme" with dummy expanding (a+b)^2 as a^2 + b^2, midwit (superimposed ekernf01 profile pic) crying and expanding (a+b)^2 as a^2 + 2ab + b^2, and wizard (superimposed Sasha Gusev profile pic) expanding E[(a+b)^2] as E[a^2] + E[b^2] because E[ab]=0.
Vanilla (Linkage Disequilibrium Score Regression) By B.navez - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=436896 Strawberry (stratified LD score regression) By Ivar Leidus - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=114685374 Cranberry (cross-trait LD score regression) By Ɱ - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=42145142 Lime (Signed LD Score Regression) By Ivar Leidus - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=97307282 Melon (Mediated Expression Score Regression) By Filo gèn' - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=68127982
(flowery font "Corsiva" with bright colors) The exotic FLAVORS of LD Score Regression Vanilla (Linkage Disequilibrium Score Regression) distinguishes whether bias is from polygenicity or population structure using GWAS summary stats Strawberry (stratified LD score regression) quantifies which genomic regions are enriched for disease risk using GWAS summary stats + functional annotations Cranberry (cross-trait LD score regression) estimates genetic correlations between traits using summary stats from multiple GWAS Lime (Signed LD Score Regression) tests whether predicted allele effects directionally align with per-allele disease risk using GWAS + deep learning Melon (Mediated Expression Score Regression) quantifies how much disease risk is mediated through gene expression using GWAS + eQTL effects
table with a rundown of perturbation prediction benchmark papers, describing each study's datasets, baselines, and major findings.
Meme showing Audrey II saying "Why don't you try ... GENE REGULATORY NETWORK INFERENCE"
Screenshot says: “OCTO-vc is a transformative tool for understanding the hidden biology driving disease progression,” said Ron Alfa, M.D., Ph.D., CEO & Co-Founder, Noetik. “Leveraging its generative capabilities, we can simulate how virtual cells behave in different tissue environments, shedding light on mechanisms that were previously inaccessible. By scaling these simulations to cohorts of patients within our dataset, we can run virtual experiments to discover patient and therapeutic biology.”
side by side images of a mundane professionally-lit home electrical panel with ~20 switches and a neon-pink-lit theater lighting panel with ~50 dimmers and a touchscreen. Captioned respectively "prokaryote transcriptional regulation" and "eukaryote transcriptional regulation".
screenshot of a paper saying "NDP kinases (NDPKs) are a textbook case of housekeeping enzymes that have been known for some 60 years to catalyze nonspecific transphosphorylation reactions between nucleoside di- and triphosphates through a conserved histidine residue. It was assumed"
Screenshot of preprint saying "To benchmark the methods, we evaluated the performance of six methods: Linear, Linear-GPT, CellOracle, GEARS, scGPT, and scFoundation (Methods). We also included a basic approach that averages gene expression across all cells within all known perturbations as the prediction of unseen perturbations (referred to as KnownAverage). The benchmarking results across the 17 datasets were summarized in Fig. 2b. Notably, the KnownAverage method consistently demonstrated some of the best overall performance across all four types of metrics."
tohsneercs fo a tsop no x yb ikoN ytraCcM tuoba eht sregnad fo derirrom efil
Screenshotted abstract of a scientific paper reads "MUCH of our present knowledge of the biological processes involved in pattern formation in Drosophila is derived from segmentation analysis1, 2. Comparatively little is known about the genetic requirement and mechanisms underlying the formation and separation of germ layers by morphogenetic movements during gastrulation3. Here we show that theDrosophila gene huckebein (hkb), a member of the gap-gene class of segmentation genes4, 5, is required for germ-layer formation at blastoderm. Absence of the hkb product, an Sp1/egr-like zinc-finger protein, causes the ectodermal and mesodermal primordia to expand at the expense of endoderm anlagen. Conversely, ectopic expression of hkb inhibits the formation of the major gastrulation fold which gives rise to the mesoderm and prevents normal segmentation in the ectoderm. Thus, hkb is necessary for endoderm development and its activity defines spatial limits within the blastoderm embryo in which the germ layers are established."
Scatterplot showing inhibition of differentiation for TF knockouts as predicted by CellOracle and also according to a literature search. X axis is granulocye-monocyte, y axis is megakaryocyte-erythrocyte. Precision at the top of each axis (fraction of large predicted scores corroborated by lit) is, by eyeball, maybe 80%.
A figure from the CENTIPEDE paper showing DNase cut site density around TF binding motifs at ChIP peaks. From plus minus 100bp on x axis moving inward, it climbs steadily. Then inside plus minus 5bp (hard to see), it plummets. Sites with less ChIP signal show weaker version of same trend.
Eight charts showing target concentration versus intensity.
Screenshot of text "based on their models and the known binding characteristics of transcription factors, Horton et al. predict that ~1300 TF's from 114 diverse species will interact with STRs in a similar manner to that observed for Pho4 and MAX.