🧬 Excited to share our new preprint! DMS chemical mapping, a key technique for studying RNA structure. Everyone assumes low DMS reactivity = Watson-Crick , high = non-WC. However, analyzing 7,500 RNA structures containing known 3D structures reveals it's not that simple. doi.org/10.1101/2024... - ThreadSky | a Reddit-style client for Bluesky

josephyesselman.bsky.social • 188 days ago

🧬 Excited to share our new preprint! DMS chemical mapping, a key technique for studying RNA structure. Everyone assumes low DMS reactivity = Watson-Crick , high = non-WC. However, analyzing 7,500 RNA structures containing known 3D structures reveals it's not that simple. https://doi.org/10.1101/2024.11.21.624766

Comments

brooklynkid53bskys.bsky.social•187 days ago

Looking at the structures in Fig1, and thinking about programs that map amount of DS vs SS (see old Wartell or Lerman papers) I would imagine a lot of bfreathin at room temp ?

there is also, I cant find it now, a well known double nonWC pair like GA/TA or something that is hyper stable

incarnatolab.bsky.social•184 days ago

Joe this is a supercool study. I was wondering, have you also looked into correlations between co-mutations for bases residing on opposite strands (basically the RING-MaP/PAIR-MaP analysis from Weeks and Mustoe)? Anyway, gunna have the guys do a journal club on your paper next week :-)

josephyesselman.bsky.social•183 days ago

Hi @incarnatolab.bsky.social thanks very much. Yes we ran this already and are processing now. There are many interesting things to checkout.

josephyesselman.bsky.social•188 days ago

To correlate DMS reactivity with RNA structure, we built a massive library of 7,500 RNA constructs containing multiple junctions with known 3D structures. Our measurements are highly reproducible (R²=0.99), span four orders of magnitude, and reveal that RNA motifs have unique DMS profiles.

Comment image

josephyesselman.bsky.social•188 days ago

Analysis of our comprehensive dataset reveals DMS reactivity exists on a continuous spectrum rather than discrete states. We observe ~10% overlap between Watson-Crick and non-Watson-Crick nucleotides, demonstrating that simple reactivity thresholds cannot reliably determine base-pairing status.

Comment image

josephyesselman.bsky.social•188 days ago

We analyzed flanking WC pairs and found structural features that determine their DMS reactivity. A-U pairs are 19-fold more reactive than G-C pairs, purine neighbors increase reactivity, and junction asymmetry correlates with higher reactivity.

Comment image

josephyesselman.bsky.social•188 days ago

Our results provide a quantitative framework for interpreting DMS reactivity patterns in RNA. This enables more sophisticated structure prediction algorithms that consider local sequence context, non-canonical interactions, and three-dimensional features - moving beyond simple base-pair predictions.

josephyesselman.bsky.social•188 days ago

Our results provide a quantitative framework for interpreting DMS reactivity patterns in RNA. This enables more sophisticated structure prediction algorithms that consider local sequence context, non-canonical interactions, and three-dimensional features - moving beyond simple base-pair predictions.

josephyesselman.bsky.social•188 days ago

Most significantly, we discover that DMS reactivity correlates strongly with atomic distances in non-canonical base pairs. These quantitative relationships demonstrate that DMS chemical mapping data encodes detailed information about RNA 3D structure.

Comment image

josephyesselman.bsky.social•188 days ago

We find that 11% of non-Watson-Crick nucleotides show protection from DMS similar to Watson-Crick pairs. This protection stems from hydrogen bonding and reduced solvent accessibility. Sequence context can alter reactivity up to 100-fold in specific non-canonical pairs.

Comment image

josephyesselman.bsky.social•188 days ago

Most significantly, we discover that DMS reactivity correlates strongly with atomic distances in non-canonical base pairs. These quantitative relationships demonstrate that DMS chemical mapping data encodes detailed information about RNA 3D structure.

Comment image

josephyesselman.bsky.social•188 days ago

Our results provide a quantitative framework for interpreting DMS reactivity patterns in RNA. This enables more sophisticated structure prediction algorithms that consider local sequence context, non-canonical interactions, and three-dimensional features - moving beyond simple base-pair predictions.

rodrigoreislab.com•187 days ago

Congratulations! Very good work! Did you probe all four nucleotides (higher pH) or A/C only? I'd be interested to know how much probing at higher pH affects structure.

josephyesselman.bsky.social•187 days ago

@rodrigo-reis.bsky.social We are doing that right now. I totally agree.

aandrzejewskaa.bsky.social•187 days ago

Congratulations! Very interesting results! 👏