When I was using helixer, I found that HelixerPost would output the precision, recall and F1 score for each seq_id.
But this confused me. I only provided the genome sequence to be predicted, so how to compare it with the label when predicting? In my understanding, it seems that there must be a reference annotation file for performance evaluation?
When I was using helixer, I found that HelixerPost would output the precision, recall and F1 score for each seq_id.
But this confused me. I only provided the genome sequence to be predicted, so how to compare it with the label when predicting? In my understanding, it seems that there must be a reference annotation file for performance evaluation?