Data and code for HybridDBRpred: improved sequence-based prediction of DNA-binding amino acids using annotations from structured complexes and disordered proteins

Install

Required predictions from other web servers

DNAgenie web server

http://biomine.cs.vcu.edu/servers/DNAgenie/

DNAgenie produces result with CSV format and names it as 'results.csv'.

DNAPred web server

http://202.119.84.36:3079/dnapred/

DNAPred produces result with CSV format and names it as 'result.txt'.

disoRDPbind web server

http://biomine.cs.vcu.edu/servers/DisoRDPbind/

disoRDPbind produces result with text format and names it as 'results.txt'.

Required third-part softwares

ASAquick

Fast neural network-based predictor of solvent accessibility

http://mamiris.com/services.html

- Modify line 12 of 'serverScript/1_genRSA.pl'  

 ```

 system "/home/ubuntu/ASAquick/bin/ASAquick ..."

 ```  

 as follows:  

 ```

 system "/yourPath/ASAquick ..."

 ```  

- Modify line 8 of 'serverScript/3_genDISO.pl'  

 ```

 $IUPRED3A_dir="/home/ubuntu/iupred3";

 ```  

 as follows:  

 ```

 $IUPRED3A_dir="/yourPath/iupred3";

 ```  

IUPred3

Prediction of Intrinsically Unstructured Proteins

https://iupred3.elte.hu/

The user needs to check the path of the IUPred3 in serverScript/3_genDISO.plx.

Required Libraries

```

pip install numpy

pip install scipy

pip install pytorch==1.12.0 (or GPU supported pytorch, refer to https://pytorch.org/ for instructions)

```

Prediction

Use command

```

$ ./genALL.pl 'jobID'

```

to run predictions.

The 'jobID' refers to the workfolder that contains the same text file with FASTA-formatted protein sequences.

For instance,

```

$ ./genALL.pl example

```

Running on GPU or CPU

If you want to use GPU, you also need to install CUDA and cuDNN; please refer to their websites for instructions.

The code has been tested on both GPU and CPU-only computer.

Features and labels data

The features data is provided in 'data/features/' folder.

The labels data is provided in 'data/labels/' folder.

Citation

Upon the usage the users are requested to use the following citation:

Jian Zhang, Sushmita Basu, Lukasz Kurgan. HybridDBRpred: improved sequence-based prediction of DNA-binding amino acids using annotations from structured complexes and disordered proteins.