Hello Bioinformatics lovers,

Tommy here.

In year 4 of my PhD, I started programming. Before that I spent four years at the bench.

I cultured cells, ran PCR and RT-PCR, did molecular cloning, ran gels and westerns, and once ran a northern blot.

I knocked down genes with shRNA before CRISPR took over.

It took me 10 years to go from wet lab biologist to bioinformatician. For much of that time I felt behind.

I kept thinking I should have learned R and Python earlier.

I was wrong about that. Those four years at the bench gave me an edge I still use every week.

What the bench taught me

When you pipette for a living, you learn how experiments fail and how they succeed.

You learn what a good control looks like because you have run the bad ones. You see batch effects creep in at the bench long before they show up in a PCA plot.

That experience is why I don’t trust a cell line name on its own. MCF7 in Lab A has different biology from MCF7 in Lab B.

Cell density, passage number, and media all change it.

The data backs this up. In 2018, Uri Ben-David and colleagues at the Broad Institute profiled 27 strains of MCF7 and tested them against 321 anticancer compounds. At least 75% of the compounds that strongly inhibited some strains had no activity at all in others.

Why this matters for your analysis

A count matrix looks clean. Every row and column traces back to cells, reagents, timing, and luck.

If you went straight into bioinformatics, you may never have seen how much of that variation starts at the bench.

Here is one habit to borrow from wet lab people. Before you analyze a dataset, ask whoever generated it for the passage number, the processing date, and who ran each batch.

Put those in your metadata. They often explain more of your PCA than the biology you set out to study.

If you’re thinking about switching

Don’t regret where you started, and don’t rush to leave the bench behind. Bring it with you. Your bench sense will catch problems that a QC report misses.

It is never too late to start. I wrote my first code four years into a PhD.

Did you start at the bench too? Hit reply and tell me which wet lab habit still shows up in your code.

(and yes, you still need to learn coding in the age of AI and AI will amplify what you know!)

Happy Learning!

Tommy aka crazyhottommy

PS:

If you want to learn Bioinformatics, there are four ways that I can help:

1. My free YouTube Chatomics channel, make sure you subscribe to it.
2. I have many resources collected on my github here.
3. I have been writing blog posts for over 10 years https://divingintogeneticsandgenomics.com/​
4. Lastly, I post daily on Linkedin https://www.linkedin.com/in/%F0%9F%8E%AF-ming-tommy-tang-40650014/recent-activity/all/ follow me for daily bioinformatics tips!

Stay awesome!