NCBI datasets is definitely an improvement over E-utils, but I wish NCBI had a REST API like ENA does. I rely on ENA every time I need to get metadata for sequences, genomes, samples, etc.
It's a separate dataset for now. Even though this project started a while ago, the metagenomes were not in IMG when I processed the data for the latest IMG/VR release
I'd love to see that :)
All of his latest tools are open source, but it still surprised me to see this. USEARCH has being maintained as closed source software for so long that I never even considered this could happen
This gets worse. Code wasn't provided to the reviewers (despite the journal's policies) and a reviewer that asked for the code and weights to be made available was ignored. twitter.com/RolandDunbra...
Demis -- I think AlphaFold3 is really exciting. As Reviewer #3, I got great results from the server. I tried hard to get @Nature to urge you to release the code but was unsuccessful. I did not get it for re-review so I don't know if you responded. So why no code? @GoogleDeepMind
This is a problem across the whole field of bioinformatics. Within the machine learning context, I think that the PlasX approach (which inspired geNomad) of weighing genomes based on how similar they are to other genomes works fine, but we probably need something more robust for the fancy models
I think we share this philosophy hahaha. I agree with you on that. It makes sense
such as linking a MTase to a motif when there's a single enzyme and motif in the genome. Or based on the different types of motif structure of the different classes of MTases.
I'm getting interested in methylation and this looks great! From what I understand, the motif and MTase detection modules are completely independent, right? Do you think it is possible to try to associate MTases to detected motifs based on the REBASE search + some additional tricks (...)
Thanks! I just did this. It only corrects the title in my profile, though. It is still screwed up everywhere else. But I can live with that :)