CLEAN is now available on BioLM: https://biolm.ai/models/clean/

This model predicts enzyme EC functions directly from sequence.

BioLMSol is now available on BioLM: https://biolm.ai/models/biolmsol/

Use this model to quickly estimate protein solubility directly from sequence.

BioLM-AbLEF is now available on BioLM: https://biolm.ai/models/biolm-ablef/

This model analyzes antibody Fv regions and provides embeddings and developability insights for heavy/light chains.

ThermoMPNN is now available on BioLM: https://biolm.ai/models/thermompnn/

Predict protein thermostability from amino acid sequence.

ESM2-StabP is now available on BioLM: https://biolm.ai/models/esm2stabp/

Predict protein stability changes directly from sequence.

We've got some new ways to manage and use your Protocol results from within Notebooks!

First off, did you know that you can have a Protocol write your results to a Volume, which you can then immediately launch in a Notebook session at the end of the run?  Here is a tutorial showing you how.

Once you are in the Jupyter Lab session, clicking on the results.jsonl (or .csv) file can open the Results Explorer, where you can browse/sort/filter results, as well as extract/download individual results to structure files as well as perform bulk operations (like extracting all structure files to a zip file for download).

We've also started adding some convenient data visualizations to the Notebook sessions.  Right click on a file to see what Viewers you can open it with, such as the sequence viewer seen in the attached screenshot.

We're working on a few more visuals as well, such as structure viewers.  Got any other ideas along those lines?  Let us know.

Hyper-MPNN is now available on BioLM: https://biolm.ai/models/hyper-mpnn/

A structure-aware model for generating or evaluating protein sequence designs.

MSA Transformer is now available on BioLM: https://biolm.ai/models/msa-transformer/

Embed or analyze multiple sequence alignments using a transformer trained on MSAs.

ThermoMPNN-D is now available on BioLM: https://biolm.ai/models/thermompnn-d/

Design thermostable protein variants from a starting sequence.

ZymCtrl is now available on BioLM: https://biolm.ai/models/zymctrl/

Predict enzyme controllability and regulatory behavior from sequence.