---
name: "gthbssmtrain"
section: "1"
platform: "linux"
package: "genomethreader"
version: "1.7.3+dfsg-10+b1"
release: "Debian 13 (Trixie)"
platforms: ["linux"]
url: "https://new.carta.tech/gthbssmtrain"
---

# gthbssmtrain(1)

train splice site model

## NAME

gthbssmtrain - train splice site model

## SYNOPSIS

**gthbssmtrain** \[option ...\] GFF3\_file

## DESCRIPTION

Create BSSM training data from annotation given in GFF3\_file.

## OPTIONS

- **-outdir**

  set name of output directory to which the training files are written default: training\_data

- **-gcdonor**

  extract training data for GC donor sites default: yes

- **-filtertype**

  set type of features to used for filtering (usually 'exon' or 'CDS') default: exon

- **-goodexoncount**

  set the minimum number of good exons a feature must have to be included into the training data default: 1

- **-cutoff**

  set the minimum score an exon must have to count towards the \`\`good exon count'' (exons without a score count as good) default: 1.00

- **-extracttype**

  set type of features to be extracted as exons (usually 'exon' or 'CDS') default: CDS

- **-seqfile**

  set the sequence file from which to take the sequences default: undefined

- **-encseq**

  set the encoded sequence indexname from which to take the sequences default: undefined

- **-seqfiles**

  set the sequence files from which to extract the features use '--' to terminate the list of sequence files

- **-matchdesc**

  search the sequence descriptions from the input files for the desired sequence IDs (in GFF3), reporting the first match default: no

- **-matchdescstart**

  exactly match the sequence descriptions from the input files for the desired sequence IDs (in GFF3) from the beginning to the first whitespace default: no

- **-usedesc**

  use sequence descriptions to map the sequence IDs (in GFF3) to actual sequence entries. If a description contains a sequence range (e.g., III:1000001..2000000), the first  part is used as sequence ID ('III') and the first range position as offset ('1000001') default: no

- **-regionmapping**

  set file containing sequence-region to sequence file mapping default: undefined

- **-seed**

  set seed for random number generator manually 0 generates a seed from the current time and the process id default: 0

- **-v**

  be verbose default: no

- **-gzip**

  write gzip compressed output files default: no

- **-bzip2**

  write bzip2 compressed output files default: no

- **-force**

  force writing to output files default: no

- **-help**

  display help and exit

- **-version**

  display version information and exit