I’ve learned the hard way that a single misplaced flag in cutadapt can turn your precious RNA-seq reads into biological confetti. My morning ritual? Coffee. htop to see if my server is crying. And grep to make sure my adapter indices didn’t cross-contaminate.
At 1:00 PM, the wet-lab team sends me an email: “Hi David, we ran the PCR. Can you just ‘quickly’ align this to the genome and find every variant associated with that rare disease? Thanks! Need it by 3 PM.” I smile. I type. I invoke the sacred magic: david bioinfo
So to my fellow Davids: keep one foot in the terminal and one foot in the literature. Validate your outliers. And for the love of all that is holy—. P.S. If you see me staring blankly at a scatter plot at 4 PM, I’m not stuck. I’m just visualizing principal components and questioning my career choices. 😉 I’ve learned the hard way that a single
I found 10,000 variants. The lab expected 5. Did I mis-call indels? Is there a batch effect? Did someone accidentally use the mouse reference genome again? (It happened once. Once.) htop to see if my server is crying
As David the bioinformatician, my real value isn’t typing fast. It’s knowing when a result is biologically plausible vs. computationally correct but nonsense .
Sometimes, I’m a plumber (unclogging data pipelines). Sometimes, a detective (finding a single SNP in 3 billion base pairs). And once a month, I’m a philosopher (arguing whether a p-value of 0.051 is really non-significant).
Why ‘rm -rf’ is scarier than a pipette tip, and other truths of digital biology. Introduction: Hello, World (of Omics)