M6AFormer
Deep-learning prediction of m6A RNA modification sites
Input
Paste transcript sequences or upload FASTA. The public workflow scans candidate adenosines and scores model windows centred on them.
Drop FASTA here
Accepts .fa, .fasta, .fa.gz, .fasta.gz up to 100 MB
Advanced options
Loading models…
0.50
Using model default until changed.
Checking CUDA availability…
Summary
Prediction summary
Total candidates scored
—
Predicted positive
—
Mean probability
—
Decision threshold
—
Predictions
Sortable table, 50 rows per page. Probability bars use the decision threshold returned by the backend.
Probability histogram
Rendered by the backend and embedded directly from /api/jobs/{job_id}/plot.png.