Nucleic acid sequences scan

Translation into Open Reading Frames (ORFs)

InterProScan 5 translates nucleotide sequences in six frames into individual ORFs using esl-translate from the Easel library. Easel is used by and distributed with the HMMER software package. The esl-translate application itself and all of its dependencies are integrated in InterProScan.

After translation, a parsing step will select the N longest ORFs and inputs them for analysis. The number number of ORFs selected depends on an interproscan property with the default value of 8. This means that analysing nucleotide sequences can take much longer than analysing protein sequences because each nucleotide sequence is translated into several protein sequences.

How can I scan nucleic acid sequences in InterProScan 5?

To scan for nucleic acid sequences you must use the flag -t n or –seqtype n

-t,--seqtype <SEQUENCE-TYPE>   Optional, the type of the input sequences (dna/rna (n) or
                               protein (p)).  The default sequence type is protein.

For example running:

./interproscan.sh -t n -i /path/to/nucleic_acid_sequences.fasta

Redundant sequences and identifiers in your FASTA file

InterProScan 5 is able to handle FASTA file entries with the same sequence, but different identifiers. For instance you have the following 2 sequences in your input file:

>sequence_1
ABC
>sequence_2
ABC

InterProScan 5 will condense these into a single sequence with two identifier cross-references in the XML output file:

<nucleotide-sequence>
        <sequence md5="e9b174d63adc63bab79c90fdbc8d1670">ABC</sequence>
        <xref id="sequence_1"/>
        <xref id="sequence_2"/>
        <orf strand="SENSE" start="1" end="3">
...

and in the GFF3 output:

##sequence-region sequence_1|sequence_2 1 3
sequence_1|sequence_2   provided_by_user    nucleic_acid    1   3
...

Entries with the same identifier and the same sequence will be merged into one.

Please note: non unique identifiers are automatically made unique by adding ‘_sequential number’ in the order of their appearance (e.g. P11111 will be P11111_1 for the first protein sequence).

Improving performance

InterProScan does not select one best ORF from the input nucleotide sequence, instead it takes the ORFs generated and select N longest ORFs and inputs them for analysis. The number selected depends on the binary.getorf.parser.filtersize property that has a default value of 8. This means analysing nucleotide sequences can take much longer than analysing protein sequences.

To improve InterProScan performance while running large nucleotide input files (> 10,000 sequences) you can:

  1. First translate your sequences externally and submit the protein sequences for interproscan analysis. This is the best approach. Besides esl-translate, other options may include getorf or transeq, both part of the EMBOSS suite of bioinformatics tools.

    If you use transeq then please use the -clean option to change STOP codon positions from '*' to 'X' because Interproscan does not accept sequences with the '*' character.

and/or…

  1. Chunk the input and then send the chunks to InterProScan. For tips on configuring the general InterProScan CPU usage see also improving performance.

Selecting the ORFs to analyse

For improved performance, Interproscan will select the longest 8 ORFs predicted for each nucleic acid sequence. This can be changed using the new “binary.getorf.parser.filtersize” setting in the interproscan.properties file

binary.getorf.parser.filtersize=8