Wei Liu

Your "pos_weights" method doesn't seem to work

1

In train_DeepCNN.py # computing weights for imbalanced go classes class_sizes = counts[args.ontology] mean_class_size = np.mean(class_sizes) pos_weights = mean_class_size / class_sizes pos_weights = np.maximum(1.0, np.minimum(10.0, pos_weights)) pos_weights = np.concatenate([pos_weights.reshape((len(pos_weights), 1)), pos_weights.reshape((len(pos_weights),...

Dataset problem

2

## Context I find some problems in the link you give about Benchmark data and Foldseek databases: http://wwwuser.gwdg.de/~compbiol/foldseek/ 1. [pdb.tar.gz](http://wwwuser.gwdg.de/~compbiol/foldseek/pdb.tar.gz) and [scp40pdb.tar.gz](http://wwwuser.gwdg.de/~compbiol/foldseek/scp40pdb.tar.gz). It seems that they are just the same....

A BUG occurs when I try to run train.py just in the default way.

2

``` iteration 1/1197 totaltrain_loss:0.695 regular loss 0.017 iteration 2/1197 totaltrain_loss:0.675 regular loss 0.017 iteration 3/1197 totaltrain_loss:0.651 regular loss 0.018 iteration 4/1197 totaltrain_loss:0.626 regular loss 0.019 iteration 5/1197 totaltrain_loss:0.601 regular loss...

Crash when using pyFAMSA

2

Hi @althonos, I am writing to express my gratitude for the updates on scoring-matrices in Version 0.4.0. The enhancements have been incredibly beneficial. I have also noticed your attempts to...