KAZUNORI YAMADA PROFILE

SPBuild

Synthetic Profile Builder


Abstract

SPBuild is a de novo profile generator based on a deep learning algorithm, LSTM. It accepts a sequence in fasta format as a sole input and generates position specific scoring matrix (PSSM).


Download

SPBuild standalone

SPBuild

Benchmark dataset

SCOP20 test dataset

SCOP20 learning dataset


Installation

You will need about 20GB memory to build the program.

$ tar zxvf spbuild-20200107.tar.gz
$ cd spbuild/code/spbuild/cc
$ ./configure
$ make
$ make install

Usage

SPBuild accepts a fasta format file of an amino acid sequence as an input file and outputs PSSM file in mtx and/or asnt format.

$ spbuild -i <fasta> [-m <mtx output file>] [-a <asnt output file>]

The output file (.asnt) can be used as an input file of PSI-BLAST search.

$ psiblast -in_pssm <asnt output file> -db <database file>

Reference

Yamada KD, Kinoshita K, De novo profile generation based on sequence context specificity with the long short-term memory network, BMC Bioinformatics, 19(1):272, 2018