search

OrensteinLab / G4mismatch

1 stars 0 forks source link

Interpreting G4mismatch #2

Open shrimonmuke0202 opened 7 months ago

shrimonmuke0202 commented 7 months ago

Hello, your work is impressive. Could you please share the source code for reproducing the interpretation of the G4 mismatch? I am interested in reproducing your analysis and would greatly appreciate access to the code.

Could you please explain how you labeled the test data reserved for chromosome 1 with '1' and '0', referring to the ground truth? After running the 'prep_data.sh' script, I obtained real values on the Y-axis instead of binary labels.

OrensteinLab commented 5 months ago

We currently do no plan on providing the interpretability code.

G4mismatch solves a regression problem. The labels are the mismatch scores calculated by the developers of G4-seq, which assigned a G4-propensity score for 15nt-long bins that server as a partition across the genome.