Arc Virtual Cell Challenge: A Primer

Arc Institute recently unveiled the Virtual Cell Challenge. Participants are required to train a model capable of predicting the effect of silencing a gene in a (partially) unseen cell type, a task they term context generalization. For ML engineers with little to no biology background, the jargon and required context can seem quite daunting. To encourage participation, we recapitulate the challenge in a form better suited to engineers from other disciplines.

Goal
Train a model to predict the effect on a cell of silencing a gene using CRISPR.

Doing things in the world of atoms is expensive, laborious and error prone. What if we could test thousands of drug candidates without ever touching a petri dish? This is the goal of the virtual cell challenge — a model (most likely a neural network) that can simulate exactly what happens to a cell when we change some parameter. Given that tightening your feedback loop is often the best way to speed up progress, a model capable of doing this accurately would have significant impact.

To train this neural network, we will need data. For the challenge, Arc has curated a dataset of ~300k single-cell RNA sequencing profiles. It may be worthwhile to revisit the Central Dogma before continuing. This essay will build off of this to provide the ~minimum biology knowledge you'll need for the challenge.

Training data

The training set consists of a sparse matrix and some associated metadata. More specifically, we have 220k cells, and for each cell we have a transcriptome. This transcriptome is a sparse row vector, where each entry is the raw count of RNA molecules (transcripts) that the corresponding gene (our column) encodes for. Of the 220k cells, ~38k are unperturbed, meaning no gene has been silenced using CRISPR. These control cells are crucial as we will see shortly.

We can see that the cell with TMSB4X silenced has a greatly reduced number of transcripts compared with the control cells.

Modelling the challenge

where: