This work is a four-year journey that would not be possible without my amazing partners @mailejim.bsky.social, Mounica Vallurupalli, and Kai Cao, and other co-authors. And thank you to my PI Fei Chen, the Golub Lab, @broadinstitute.org, @harvardmed.bsky.social, for your support! (10/10)

Our work shows that RNA splicing can be harnessed as a new modality for cell type-specific gene regulation, unlocking new possibilities for gene therapy and precision medicine. It also highlights how high-quality experimental data powers AI models. (9/10)

We further applied SPICE to design sequences that specifically target cancer cells carrying splicing factor mutations. We identified synthetic sequences that are selectively spliced only in cells with RBM5 or RBM10 mutations, which are commonly seen in lung cancers. (8/10)

To go beyond prediction, we built Melange, a generative model that designs new sequences with programmed cell type-specific splicing. We experimentally validated that Melange can design sequences that splice only in neural-lineage cells, and not cells from other lineages! (7/10)

Using this dataset, we built Soma, a deep learning model that predicts how any RNA sequence will splice across cell types.

We drew inspiration from ChromBPNet (@anshulkundaje.bsky.social) to encode cell type-specific information directly from gene expression data. (6/10)

We profiled 46,000+ sequences across 43 cell lines spanning 10 lineages – the most diverse set of cell types ever tested in an MPRA (>10x more than previous studies) – and uncovered widespread cell type–specific splicing events. (5/10)

Good ML models require high-quality experimental data, but large-scale, cell type-resolved data on RNA splicing is limited. Hence, we built SPICE (Splicing Proportions In Cell types), an integrated experimental + AI framework for the generative design of cell type-specific RNA sequences. (4/10)

Beyond existing cell type targeting strategies like promoters, enhancers, or viral vectors, we asked: can we design RNA sequences that use alternative splicing as an underexplored biological process to control cell type-specific gene expression? (3/10)

Every cell in your body contains the same DNA, but different cell types – like neurons or cancer cells – perform vastly different functions. Being able to turn genes on or off in specific cell types is key for understanding biology and for building precise, safe gene therapies. (2/10)

Nah you're still a cow! Mooooo.

Here’s my spreadsheet of starter packs (>80!) related to broadly ‘mechanistic biology’ plus some intriguing extras

Complete with collective nouns

I’ve been tracking these but now can’t keep up

Hope it’s helpful

2/2

docs.google.com/spreadsheets...