Understanding garnish_antigens output table

[antaralabiba97]

Hi,

So I have attached my output for the function "garnish_antigens" with the default parameters set in the user manual. I was trying to understand the meaning of all the columns as the manual doesn't have all the outputted columns named.

I found this the case for the other functions too where extra columns are added even with the test data...

Mainly from the attached file, I was wondering how the terms included in the "Recognition_Features" is determined and how the "min_DAI" is calculated and the meaning?

Also, a foreigness_score of 1 indicated greater homology to immunogenic peptides which I assume is good for a predictive neoantigen. However, I would then assume the dissimilarity score to be near 1 also? The last thing is I am not sure what the Ensemble_score is out of because in the results for garnish_affinity the values vary considerably, therefore, I am wondering what type of score is deemed good for an immunogenic peptide using your algorithm.

Thank you in advance!

garnish_antigen_output.csv

[andrewrech]

"Recognition_Features" is created here from output columns:

https://github.com/andrewrech/antigen.garnish/blob/fb9458b462c2f6d164b301eb14503f6979f8910e/R/antigen.garnish_io.R#L296-L303

min_DAI is documented in ?garnish_affinity: "Minimum of value of BLAST_A or DAI values, to provide the most conservative proteome-wide estimate of differential binding between input and wildtype matches."

Dissimilarity and foreigness are defined at length in the associated paper linked in the README, as well as their performance for perdicting immunogencity. I suggest looking at the empiric data we provide in this manuscript. A "good" score is highly contextual.

Richman LP, Vonderheide RH, and Rech AJ. Neoantigen dissimilarity to the self-proteome predicts immunogenicity and response to immune checkpoint blockade. Cell Systems. 2019.