Skip to content

InterProScan data import

Patrick Durand edited this page Feb 5, 2021 · 1 revision

About Interproscan in BLAST Viewer

Starting with BLAST Viewer release 5.5.0, it is possible to import XML or GFF3 data files produced by InterProScan.

Memory (RAM) issue

Importing large amount of BLAST and InterProScan result files may require large amount of RAM.

So, you should have to start BlastViewer with appropriate RAM allocation through the use of the Java command line. More here.

How to import IPRscan data files?

First of all, import BLAST results into BLAST Viewer:

Such results should contain same queries as the ones provided to InterProScan to predict functional domains. Indeed, BLAST Viewer import these domains using query IDs.

As soon as BLAST results have been imported, one can see on the summary overview panel that no classification data is available:

Now, switch to the "Query/hit table":

And click on the "Import IPRscan" button:

Select all IPRscan data files to import (here GFF3 format):

BLAST Viewer opens a new frame containing the merge of BLASTx results and InterProScan predictions:

By switching again to the "Query/hit table" panel, one can see that "Query classification" column is filled in with InterProScan data:

Using "Classifications" button (bottom right of the Query/hit table), you can control types of classification displayed in that table.

GFF3 or XML Interproscan data files?

GFF3 data import feature was added first in BLAST Viewer. However, there is a bad issue with that format: GO and IPR description fields are not reported.

So, XML IPRscan data format should be the prefered one since it contains more information that GFF3.

Review IPRscan predictions with details

BLAST Viewer let you review the details of InterProScan predictions as follows.

From the "Query/hit table" panel, select a row displaying a query annotated with some classifications, then click on "View IPRscan" button:

You will get a dedicated viewer mapping IPRscan predictions on the query sequence:

Link to data provider web sites

It is worth noting that the Feature table Viewer (see previous screenshot) displays a little Earth icon next to each domains: click on that icon to get full details directly on data provider web sites.