Step 1: Download the stand-alone package file to your runnable linux device with python 3.5 and above environment. Unzip the file using the following commands:
> unzip IDP_LM.zip
> cd IDP_LM
Download model file here1 and copy it to "/IDP_LM/protTrans/prot_t5_xl_uniref50/".
Download model file here2 and copy it to "/IDP_LM/protTrans/prot_bert/".
Step 3: Create and activate the required environment of IDP-LM using the following commands:
> conda env create -f IDP_LM/torch.yml
> conda activate torch
Step 4: The sequences to be predicted were uploaded in fasta format file "examples.txt", and using the following commands to generate the intrinsic disorder prediction results:
> sh run.sh examples.txt disorder
Wait until the program completed, and you can find the final prediction results:"/IDP_LM/temp/results/IDR_results.txt"
Step 5: Using the following commands to generate four common disorder function prediction results (disorder protein binding, disorder DNA binding, disorder RNA binding, disorder flexible linker):
> sh run.sh examples.txt function pb
> sh run.sh examples.txt function db
> sh run.sh examples.txt function rb
> sh run.sh examples.txt function linker
The corresponding result files are avaliable at "/IDP_LM/temp/results/"