←
Kepler Mears107 posts
Kepler Mears avatar

Kepler Mears

@KeplerMears

I much prefer the sharpest of criticism of a single intelligent man to the thoughtless approval of the masses - Johannes Kepler

◷ Joined April 2014▧ 30 Photos and videos
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

Arya (my best friend) does not have social media, he asked me to tweet this for him. But if you want to chat about science (or offer him a job) you can contact him at arya.casa/contact. He’ll (probably) be defending soon!

Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

Some ~caveats~ 🔹 We haven't proven any deletion was beneficial 🔹 No functional deletion-born fusion yet - selection analyses show mainly diversifying selection 🔹 Our method misses fusions with terminal mutations (requires exact 54bp matches)

Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

The "deletional bias" in bacterial genomes has long been framed as destructive - the pruning away of genetic material. Our findings suggest it might also be creative: deletions don't just remove genes, they can also build new ones!

Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

The prefix-suffix approach isn't just for fusions. We also found: 🔹 Internal deletions 🔹 Repeat prophage insertions 🔹 Variable gene cargo in mobile elements It's a general tool for surveying structural variation at scale!

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

But here's the CATCH: detection depends on sampling depth. When we queried well-sampled protein families we found significantly more fusions than from uniformly sampled families. We're only seeing the tip of the iceberg. As databases grow, we expect to find more.

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

Applying the prefix-suffix approach to type strains from 5 commonly studied bacteria (E. coli, N. gonorr., M. tubercul., Strep. pneumo., C. jejuni) we find deletion-born fusions across the bacterial Tree of Life:

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

How common is this across the bacterial tree of life? We developed a new method to query 2.4M+ genomes (thanks to @zamiqbal's AllTheBacteria!): look for the “prefix” and “suffix” k-mer of an input gene at variable distances. If the distance changes → structural variation.

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

It's not just in the lab. During the M. tuberculosis-M. bovis split, we found 2 deletion-born fusions: 🔹mlaE-htpX 🔹acrR-glcD These arose during speciation and persist in natural populations today.

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

We found one in the Lenski LTEE: a 57.5 kb deletion in E. coli in the Ara+1 lineage fused yjcO (unknown function) to lysU (lysyl-tRNA synthetase). Sequencing shows it's transcribed, translated, and swept to fixation in ~500 generations.

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

Most new genes die before they do anything useful, lost to genome streamlining and drift. But deletion-born fusions are different! They hitchhike on beneficial deletions, getting a "free ride" to high frequency. This buys them TIME to evolve function.

Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

Here's the idea: when a deletion is beneficial (reducing genome size, removing costly genes), it can bring two distant gene fragments together. The result? A brand new fusion gene - created entirely by accident.

Post attachment
Kepler Mears avatar
Kepler Mears@KeplerMears·Jan 6

Duplication is constrained by existing templates. Overprinting produces disordered proteins & must preserve the original gene. HGT just imports genes born elsewhere. What if deletions - usually viewed as destructive - could also create?