We have been training our binary decision tree on psms rather than peptides. PSMs for the same peptide are not independent measurements. Therefore, they are not a good choice for ML training applications. This PR uses peptides instead and then applies that model to the PSMs.
Also, there are some efficiency and orangizational improvements.
We have been training our binary decision tree on psms rather than peptides. PSMs for the same peptide are not independent measurements. Therefore, they are not a good choice for ML training applications. This PR uses peptides instead and then applies that model to the PSMs.
Also, there are some efficiency and orangizational improvements.