Edit in GitHub

lncScore is an alignment-free tool to detect lncRNAs from assembled transcripts, and it also can be used to calculate the coding potential.

Download the latest version of lncScore. Version: 1.0.0. Last Modified: 2016-1-31

Download here

Authors: Jian Zhao (zhaojian@nuaa.edu.cn or zhao_doctor@hotmail.com)

You can also modify, edit GFusion, and give us your suggestions and comments in GitHub, or by email.


Usage: lncScore.py [options]


-h, --help show this help message and exit

-f input files, --file=input files

enter transcripts in .bed or .fasta format: if this is

a .bed format file, '-r' must be specified; if this is

a .fasta format file, ignore the '-r'.

-g gtf file name, --gtf=gtf file name

please enter your gtf files

-o output files, --out=output files

assign your output file

-p prallel numbers, --parallel=prallel numbers

please enter your specified speed ratio

-x hexamer matrix, --hex=hexamer matrix

Prebuilt hexamer frequency table (Human, Mouse,Fly,

Zebrafish, C. elegans, Sheep and Rat). Run

'make_hexamer_tab.py' to make this table out of your

own training dataset.

-t training dataset, --train=training dataset

Please enter your specified training dataset

-r reference genome files, --ref=reference genome files

Reference genome sequences in FASTA format. Ignore

this option if sequences file was provided to '-f'.

Reference genome file will be indexed automatically

(produce *.fa file along with the original *.bed file

within the same directory) if hasn't been done.


  • If the input is a FASTA format file, then make sure that the description line (begins with a ">" symbol) only contains the transcript ID, for example: "ENST00000379410.7", which should be found in the GTF format file.
  • If the input file in *.bed format, then an additional python package named pysam was required to be installed first.
  • Example:

    python lncScore.py -f test/human_test.fa -g test/human_test.gtf -o result -p 1 -x dat/Human_Hexamer.tsv -t dat/Human_training.dat

    Software Prerequisites:

    The following two software should be installed in your cluster or computer before running the lncScore.py

  • Perl (>=5.10.1)
  • Python (>= 2.7),
  • The scikit-learn model

    Output file format:

    Columns (from left to right):

    1. Transcript id;

    2. Index;

    3. Coding score.

    For example:

    Transcript_id Index Coding_score

    ENST00000342066.7 coding 0.999999999941

    ENST00000327044.6 coding 1.0

    ENST00000338591.7 coding 0.999999999998

    ENST00000562538.1 noncoding 0.0950315486986

    ENST00000415166.1 noncoding 0.0264167916398

    ENST00000565955.1 noncoding 0.0149329291793

    Copyright (C) 2016 Xiaofeng Song.