Our paper with @eternagame, 5 AI teams, and cryo-EM collabs on de novo design of #RNA pseudoknots with #deeplearning is out in @ScienceMagazine ! Comes with a provocative commentary on RNA structure prediction by @TonyMustoe lab! https://t.co/UuwyK18qdU https://t.co/jtNjJPe0Ir
New preprint: https://t.co/P907xjCohp
I'm excited to share our latest work from a long-standing collaboration with Jessica Disney, Yanjun Li, and Chenglong Li on using AI/ML approaches to identify and understand small molecule-RNA molecular recognition. We are grateful to the students who drove these studies—Amirhossein Taghavi and Riddhish Thakare—with support from Jielei Wang.
Here, we introduce SMARTPocket, a geometric deep learning approach to identify ligandable pockets in RNA. The key idea is transfer learning: SMARTPocket first learns principles of molecular recognition from more than 110,000 protein interfaces with small molecules and then applies that knowledge to RNA, where far less training data exist. We validated this approach across a wide variety of benchmarks, where it outperforms existing methods. The model generalizes to apo RNA structures when conformational changes are modest, identifies cryptic ligandable pockets, and recapitulates experimentally validated binding sites in the SARS-CoV-2 frameshifting element and an RNA aptamer evolved to bind small molecules.
This paper is part of a larger vision in our laboratory that started with Inforna (Nature Chemical Biology, 10, 291–297 (2014)) and other earlier work. The goal is to build computational and experimental approaches to accurately identify RNA-binding and ultimately bio-functional small molecules. More work is needed for these approaches to fulfill their promise.
We will roll out additional augmentations to this long-standing program this summer, so stay tuned for more. Thanks for reading—RNA FOREVER!
As a Guest Editor for @JoVEJournal, I am organizing a Collection titled “Frontiers-in-rna-structural-biology-mechanisms-methods-and-therapeutic-potential.” Interested in contributing? Please submit an abstract here: https://t.co/KoKVhMZtF3
Today we’re releasing an alpha version of RibonanzaNet2. Rnet2 is a 100M-parameter foundation model for #RNA structure, trained on chemical mapping profiles for 30M RNAs with complex structure. https://t.co/aN72ZoMd9e (1/N)
How sturdy are base pairs near the ends of #RNA hairpins? Perhaps not as stable as one might think, esp. when there are bulges near and far. Led by @CharlesBNader in @NatureComms, we describe complete structures of #HIV#TAR RNA and its dynamic excursions.https://t.co/2vvKK1CQaC
Do you love RNA biology and RNA structure? Join us at Atomic AI to shape the future of AI-driven drug discovery targeting RNA! We’re hiring for a computational RNA expert - check out this job posting here: https://t.co/l85R3qSKCF
To give an idea of the rate of progress in science, in the last 2 days @Nature (alone) has published pre-clinical discoveries of:
1) New antibiotics
2) A new obesity medication
3) A new pain medication
4) New atherosclerosis medications
Our revised manuscript explains how Guo et al. concluded PRC2 is not an RNA binding protein. They retained unmappable reads, selectively removed read duplicates from the mappable fraction, and applied an arbitrary cutoff for PRC2 enrichment in RNA.
https://t.co/SbN5bJJX5J
Same goes for science I echo his voice for PHds just need to do something so they feel stable after earning the degree …we as society are failing my heart breaks when the students tell they don’t have jobs after getting the degree ..
https://t.co/66mpUfCGK5
Last year Purba moving to Germany for her PhD than Ajit defending his thesis successfully and this year Shalini moving to Austria for PhD . It is self satisfying as a Mentor to see them doing good .. My wish is they all do great in their life all the best ..very proud movement.
Time flies and I have been engaged in RNA crystal structure research for nearly 20 years, finally arriving at an opportunity to make methodological contributions. This study combines the concepts of flexibility and overcoming rigidity. https://t.co/aLqOg1ye9j
When asked to draw a scientist, school-age kids in the United States are increasingly sketching women, according to a study from 2018.
Read more on #STEMSTEAMDay: https://t.co/xtOAX8sof0