How to use MSA Transformer Online

Use MSA Transformer online for alignment-conditioned mutation scoring and protein sequence generation.

MSA Transformer, also known as ESM-MSA, learns protein-family constraints from multiple sequence alignments. This service uses the ESM-MSA-1b checkpoint described by Rao et al. (2021). Optional sequence generation uses Sean Johnson's protein Gibbs sampler, building on Johnson et al. (2021) and its later enzyme generation study.

How MSA Transformer Works

Upload an aligned FASTA or A3M file with the query first. At least two distinct sequences are required. Alignments may contain up to 512 columns and 4096 rows. A3M insertions and exact duplicate rows are removed, and up to 31 evenly spaced context sequences are retained.

Choose Mutation Scoring (without Gibbs) to mask each non-gap query residue independently. The CSV reports every canonical substitution, its natural-log likelihood ratio relative to wild type, and a probability normalized over the 20 canonical residues. Query Position counts ungapped residues; Alignment Column counts columns in the insertion-stripped alignment.

Choose Sequence Generation (Gibbs) to resample only the query while holding context rows fixed. Set the number of sequences, complete sampling passes, burn-in passes, and top-k restriction after burn-in. Each pass resamples all query columns in random groups, including gaps. CSV outputs preserve aligned sequences; FASTA outputs remove gaps. Identity to Query is the fraction of non-gap query positions matched by the design. Neither model preferences nor generated sequences establish experimental function.

What is Neurosnap?

Neurosnap is the leading platform for bioinformatics and computational science focused on expanding access to powerful modeling and simulation tools. Because many state-of-the-art machine learning systems remain complex to install, configure, and scale, Neurosnap offers a clean, browser-based workspace that removes the burden of infrastructure management, dependency conflicts, and command-line tooling.

Built for biologists, chemists, and cross-disciplinary scientists, the platform enables advanced computational workflows without requiring expertise in software engineering or cloud architecture. Researchers can launch analyses through an intuitive interface, connect programmatically through a comprehensive API, and rely on automated resource management to scale workloads efficiently. By taking care of the underlying compute and operational complexity, Neurosnap allows teams to devote their energy to scientific progress and faster iteration. Security and data protection remain foundational principles, with clear safeguards outlined in our Terms of Use and Privacy Policy to ensure your work stays protected.

Advancing Discovery with MSA Transformer on Neurosnap

Using MSA Transformer on Neurosnap could drastically accelerate alignment-conditioned mutation analysis and sequence design with optional Gibbs sampling.

Family context: Condition predictions on an uploaded alignment.
Two workflows: Score substitutions without Gibbs or generate query variants with Gibbs.
Reusable results: Export substitution scores as CSV or designs as CSV and FASTA.

How to Use MSA Transformer on Neurosnap

To harness the capabilities of MSA Transformer, researchers can follow this streamlined workflow within Neurosnap:

Access Neurosnap: Start by logging in to the Neurosnap website.
Select Tool: From the list of available tools, choose MSA Transformer.
Provide Inputs: Provide all the inputs specified within the submission panel and optionally configure the tool as desired.
Run Tool: Submit the MSA Transformer job and Neurosnap will execute it in the cloud, automatically notifying you as soon as your results are ready.
Review Output: Explore your results through rich visualizations, including figures, plots, and interactive views designed to help you analyze findings with clarity and confidence.

Citations

Please cite the original work when using MSA Transformer in publications or research outputs.

Rao, R. M. et al. MSA Transformer. Proceedings of ICML, PMLR 139:8844–8856 (2021). https://proceedings.mlr.press/v139/rao21a.html

Johnson, S. R., Monaco, S., Massie, K. & Syed, Z. Generating novel protein sequences using Gibbs sampling of masked language models. bioRxiv (2021). https://doi.org/10.1101/2021.01.26.428322

Johnson, S. R. et al. Computational scoring and experimental evaluation of enzymes generated by neural networks. Nature Biotechnology (published online 2024). https://doi.org/10.1038/s41587-024-02214-2

Neurosnap Inc. (2022). Neurosnap: An online platform for computational biology and chemistry. Available at: https://neurosnap.ai/

Similar Services

Explore related tools that support similar research workflows:

ESMC Mutation Scoring

ProSST Mutation Effect Prediction

BindFilter

Efficient Evolution

Proudly supporting 50,000+ scientists worldwide, including 7,000+ leading biotech and global biopharma organizations.