Dataset understanding

[dberma15]

I'm looking to study the effect of mutations on antibody escape and I was wondering if you could let me know if I'm looking at the ideal data for this.

1. Is the file results/escape_scores/scores.csv the ideal file for this? I have tried looking through a lot of the repositories under the Bloom Lab and it is the only one I can find that seems to have multiple mutations from the wild type. For example, in the column "aa_substitutions", I've found values like "S19R G51K T100N P149L G166R", which indicates multiple mutations.
2. If this is the right dataset, I assume score is the correct column to use, correct?
3. If this isn't the right dataset, what dataset should I use?

[tylernstarr]

If you're interested in the effects of (sparsely sampled) multiple-mutant genotypes on antibody escape, that would be the proper file. The "best" final data file that basically collapses phenotypic scores to the per-mutation level is here, where the column mut_escape is your metric of interest.

This repository has escape for the first dozen or so antibodies we've mapped, but we and others (including one group that does experiments at massive throughput) have mapped many additional antibodies. @jbloom has put together repositories that compile across all of these various antibody escape datasets, and he can probably best point you to the most recent data compilation table (I have my idea where it is but might not link the best option).

[dberma15]

Out of curiosity, how does the dataset you referenced compare to this the dataset here?

[tylernstarr]

The latter dataset as described in its repository is a compilation of many different antibody escape datasets across multiple publications, and so is likely a better dataset to use for general trends in antibody escape