GIGEASA-BS:
To compute binding site derived features for GIGEASA model for a large number of proteins from PSSM data
|
This server
takes upto 1000 PSSM files as input (maximum zipped file upload is 10MB) and returns GIGEASA binding sites features based on the predictions for DNA, RNA, PPI, ATP and carbohydrates binding sites.
These features are described in Ahmad et al. 2017 (under review).
To read notes on how to create a PSSM file and data for upload, click Here .
All pssm files must be in a PSIBLAST output format. An example zipped file for upload is available Here .
|
To start running GIGEASA-BS, simply upload a zipped file containing PSIBLAST calculated PSSM files (Maximum file size is 10MB). If you have more than this size of data, please upload them as multiple files
Note: Upload and feature computing time is estimated to be
around 20 seconds per pssm (~200 residue sequence; benchmarked on uploads of 100 proteins each), during usual server loads. Please be patient if your uploading large number of PSSMs.
Precomputed data sets of these features can be downloaded for UniProt entries of three species, as follows:
Human proteome BS based features
Mouse proteome BS based features
Arabidopsis Thaliana proteome BS based features
Whole proteome predictions for these species from different prediction models are available as follows
Human proteome predictions:
Models trained on UniProt FT definition of DBP
Models trained on GO terms definition of DBP
Models trained on PDB complex based definition of DBP
Models trained on Pfam complex based definition of DBP
Mouse proteome predictions:
Models trained on UniProt FT definition of DBP
Models trained on GO terms definition of DBP
Models trained on PDB complex based definition of DBP
Models trained on Pfam complex based definition of DBP
Arabidopsis Thaliana Predictions:
Models trained on UniProt FT definition of DBP
Models trained on GO terms definition of DBP
Models trained on PDB complex based definition of DBP
Models trained on Pfam complex based definition of DBP