←
Franz Baumdicker213 posts
Franz Baumdicker avatar

Franz Baumdicker

@fbaumdicker

Mathematical and Computational Population Genetics Microbial Evolution | Machine Learning in PopGen | CRISPR-Cas | Pan-genomes | HGT 🦠 ➔ 🧬 ➔ 💻 ➔ 🤔

◷ Joined November 2018▧ 11 Photos and videos
Nora Pyenson avatar
⟳ Franz Baumdicker retweeted
Nora Pyenson@norapyenson·12 Dec 2024

Is it “winner-takes all” when the simplest living things compete? Check out my fresh publication on phage coexistence in Science and a thread below🧵 science.org/doi/10.1126/scie…

Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

It is crucial to use reconstructed arrays to see these deletion patterns, as the position of spacers shifts toward the trailer end over time. I will post more details about our reconstruction tool SpacerPlacer later. But you can find the tool here: buff.ly/40QdDId

Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

Finally, across multiple CRISPR-Cas types and genera, we found no significant difference in deletion rates or lengths, suggesting that spacer deletions are likely not modulated by Cas proteins.

Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

CRISPR spacer deletions may be more common at the leader due to a) mutations of the last repeat b) recently acquired spacers lacking a long-term benefit c) benefits of changing the spacers at the first most expressed positions d) (reconstruction errors) e) all of the above

Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

The spacer deletions at the very first and last positions of CRISPR arrays deviate from a symmetric boundary effect. The reconstructed deletion rate is higher at the first positions compared to the losses at the last positions. This might have multiple reasons ...

Post attachment
Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

Besides the boundary effect, we observe no change of deletion frequencies along arrays. They evolve surely non-neutrally, but our results hint at either weak selection pressures to delete specific spacers or an equally strong selection along the array to maintain spacers. But,...

Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

Consistent with this mechanism, we observe a “boundary effect”: Spacer deletions are less frequent near the leader and trailer ends of the array, potentially due to fewer neighboring repeats for alignment, while the middle of the array sees the most deletions.

Post attachment
Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

Interestingly, besides the loss of a single spacer, deletions occur often in blocks, involving an average of 2.7 spacers per event. This block deletion aligns with theories of replication slippage, where nearby repeats misalign, leading to multiple spacer deletions.

Post attachment
Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

Spacer deletions in CRISPR arrays follow fascinating patterns. Using our tool SpacerPlacer, we found that deletion events are 374x more frequent than single-nucleotide mutations, making immunity loss by spacer loss much more likely than by mutations.

Franz Baumdicker avatar
Franz Baumdicker@fbaumdicker·20 Nov 2024

CRISPR arrays store "immune memories" of past invaders as spacers. But these memories aren’t permanent— frequent spacer deletions reshape the array. We investigated the spacer deletion dynamics with our new reconstruction tool and learned a lot... 🧵 buff.ly/4fS4Iu3