Tutorial

My first blog post in a long time! I wanted an AI assistant that knew me. Knew my research. Knew my content calendar. Knew that when I say “write a thread about batch effects,” I mean single-cell RNA-seq batch effects with Harmony and Seurat code, for an audience of computational biologists who already know what a UMAP is. So I built Helix. An OpenClaw agent running on a Mac Mini in my home office in Boston, connected to my Google Calendar, my content docs, and a deep memory system that learns my preferences over time.

I started to learn bioinformatics because I needed to analyze public ChIP-seq data in 2012. That’s how I got to know Shirley Liu’s lab at Dana-Farber Cancer Institute. And God knows that I would join her group in 2020 for a staff scientist position to lead the CIDC bioinformatic project. I witnessed the development of many groundbreaking computational tools for genomics in Shirley’s lab. One tool that I found particularly elegant was BETA (Binding and Expression Target Analysis), developed by Su Wang and published in Nature Protocols in 2013.

During my work with single-cell RNA-seq data, I’ve often encountered confusion about PCA and specifically when to use the center and scale arguments in R’s prcomp() function. While tools like Seurat’s RunPCA() abstract away these details, understanding what happens under the hood is crucial for proper analysis and troubleshooting. In this post, I’ll show you exactly what center and scale do, why they matter, and what happens when you get them wrong.