revisited a past project and made a visual guide to G-Computation estimation for causal inference🎨 G-comp is also called the g-formula or g-standardization. pdf/jpg on Github👉 github.com/kathoffman/cau… older TMLE and superlearning guides linked below. IPW and AIPW coming soon!

Excited to reconstruct and analyze *Dynamic* Gene Regulatory Networks along single cell RNA+ATAC trajectories with Lingfei Wang and Nikolaos Trasanidis ! Check out our new method Dictys ( <δίκτυον ) biorxiv.org/content/10.110…

I am seeking a postdoc (jointly supervised by Dr. Haibo Zhou), focusing on environmental health. US citizen or green card is required, the expected start date is no later than May 2023.

Happy to share our latest work (w/ Natasa Tagasovska, Stephen Ra, Kyunghyun Cho, Jonathan Pritchard and Aviv Regev) about learning causal representations of single-cell perturbation data sets arxiv.org/pdf/2211.03553….

Really useful package! And it was applied in the bootRanges vignettes. It’s a wonderful collaboration!

Excited to share this work, collab with Michael Love et al!! Diffsig can test the association btwn mutational signatures and “risk factors” which can be any variables of interest – continuous or categorical – and multiple risk factors can be tested at once: biorxiv.org/content/10.110…

🎉🧹 The #tidyomics ecosystem is official! Into #omic data analysis? Spanning #Seurat Bioconductor #SCE, #SE, #GRanges, #Citometry? Now, just use #tidyverse ! Co-led with Michael Love and Prof Tony Papenfuss WEHI (Walter and Eliza Hall Institute) #singlecell The preprint biorxiv.org/cgi/content/sh…

Big shoutout to my amazing student Wancen Mu and colleague Yun Li for their hard work on this project! 🌐 Don't miss our detailed findings in our paper: arxiv.org/abs/2310.17023 #TeamWork

Join us at the WiML at #NeurIPS2023! 🙌 My students Jiawen Chennn, Wancen Mu, and Christine Park are presenting three posters on diverse topics: Gradients in Spatial Transcriptomics, Lead-Lag Effect in Time Series, and Network Curvature for community detection.

Thrilled to witness our work reach publication! Many thanks to Naim Rashid for the invaluable guidance!

Google presents Does Fine-Tuning LLMs on New Knowledge Encourage Hallucinations? Highlights the risk in introducing new factual knowledge through fine-tuning, which leads to hallucinations arxiv.org/abs/2405.05904

20 epochs for finetuning a LLM... this is fascinating