Run workflows using AF3

MSAs generated with ColabFold and predictions ran with AlphaFold3.

1. Custom prediction

Run modelling of a custom protein or complex. In this case, MSAs are generated using ColabFold server. Not suitable for RNA modelling.

bash run_custom.sh example_1.fa

Input file:

>Complex1|Q92759_A:Q92759_B:dna_for:dna_rev
FIRSTPROTEIN:SECONDPROTEIN:dna|ACGT:dna|TGCA

For more information on how to format an input file, visit this page.

2. Batch prediction

bash run_batch.sh [options] example_2.fa

Input file:

>complex_1|protA:protB
FIRSTPROTEIN:SECONDPROTEIN
>complex_2|protA:protB
THIRDPROTEIN:FOURTHPROTEIN
>complex_3|protA:protB
FIFTHPROTEIN:SEVENTHPROTEIN

3. DNA binding prediction

Run modelling with standard DNA probes in batch. Input fasta file should contain homodimers. The results of this computation can be used to infer protein-DNA interaction.

bash run_batch.sh -d [options] example_3.fa

Input file:

>complex_1|protA:protB
FIRSTPROTEIN:FIRSTPROTEIN
>complex_2|protA:protB
SECONDPROTEIN:SECONDPROTEIN
>complex_3|protA:protB
THIRDPROTEIN:THIRDPROTEIN