Skip to content



CD-HIT can be used for clustering large sequence sets or removing identical or highly similar sequences from a sequence set. CD-HIT is often used as a tool to produce a non-redundant sequence set for further analysis of a large sequence set. CD-HIT recognizes fasta and fastq sequence formats.


Free to use and open source under GNU GPLv2.


Version on CSC's Servers

Puhti: 4.8.1


The setup command for CD-HIT in Puhti cluster is:

module load biokit

After the setup command, the server recognizes CD-HIT commands. The CD-HIT package has many programs. The most notable are:

Program Description
cd-hit Clustering and redundance removal tool for protein sequences
cd-hit-est Clustering and redundance removal tool for nucleic acid sequences (only for sequences that do not contain introns)
cd-hit-2d Tool to compare two protein sequence sets
cd-hit-est-2d Tool to compare two nucleic sequence sets
cd-hit-454 A program to identify artificial duplicates from raw 454 sequencing reads
cd-hit Cluster peptide sequences
psi-cd-hit Cluster proteins at less than 40% cutoff
cd-hit-lap Identify overlapping reads
cd-hit-dup Identify duplicates from single or paired Illumina reads
cd-hit-454 Identify duplicates from 454 reads
h-cd-hit Hierarchical clustering

A full list of programs can be found in the CD-HIT user guide.

You can list the command line options of CD-HIT programs by using option -help. For example:

cd-hit -help

A simple analysis for a protein sequence set can be done for example with command:

cd-hit -i my_proteins.fasta -o reduced_set.fasta -c 0.95

The sample command above produces two result files:

  • reduced_set.fasta contains a pruned sequence set. In this case, if two sequences are more than 95% identical, only the longer one is included to the results.
  • reduced_set.fasta.clstr contains information about the clustering of the sequences that share higher similarity than the give threshold value (in this case 95%).



Last edited Mon Mar 29 2021