Hello Sushil,
I can help you with setting up the config file. The gff3 and bams are outside the scope of majiq itself.
For the bam files, these are aligned experiments after using an aligner like STAR (
https://github.com/alexdobin/STAR). You'll need the original data from the sequencer, and you'll likely also want to trim and preprocess it first.
For GFF3 this will be the reference annotation that you trust as a baseline. It's important to make sure that the gene names here match whatever the sequencer provides in the previous step. A common reference to use is the ensenbl hg38 gff3 for human data, as an example.
Please let us know if you have any technical issues with majiq. If you have your experiments ready, I can certainly help you create an appropriate config file.
Thanks~