Fast, accurate and versatile sequence classification system

Computer Science & Engineering Department
University of California
Riverside, CA 92532






  The source code of CLARK is distributed under the GPL License. CLARK and its variants (such as CLARK-S) are a free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
CLARK and its variants are distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details at
Copyright 2013-2019, Rachid Ounit ( at


•   Version (July 2019):
What is new?
The script generating summary tables has been extended to include a third table: "TableSummary_HitCount" providing a comparison of reads count for each species detected across the CLARK reports.
This table sorts the species per frequency of presence across reports, which is ideal for comparison microbiome profile across samples (when each report corresponds a specific sample/metagenome).
Code improvements and bug fixes are included.

Download the version

The zipped package contains C++ files and shell script, designed for 64-bit OS, and executable on the latest Linux and Mac systems. Do you see a missing feature that should be included in the CLARK package? You have any questions? We welcome users to share their comments/feedback/suggestions in the GoogleGroup of CLARK users.

Also, for your convenience, the Frequently Asked Questions regarding CLARK's usage are available in a separate tab.


  CLARK is being supported, so feel free to report any bugs, error messages or unexpected behavior by email at at or via the Google Group of the the CLARK users. We guarantee a response within 72 hours (while so far our average response time is less than 24 hours) to any bug/error report that follows the guidelines below:

• Rerun CLARK (or the related variant) with all conditions that created the bug to make sure the bug is reproducible.
• Check whether the bug is persistent with other sample files (e.g., the publicly available datasets we used in our peer-reviewed papers) or only specific to your input data.
• If the bug is reproducible then please send an email to at or create a post at the Google Group of the CLARK users with:

1) Your contact information (at least you full name and email address) so we can contact you as soon as the fix is available;
2) All input data (targets and objects) that created the bug if the bug/error is not persistent with other samples (Create a folder containing these data and zip it) as attached files (especially for small size files). This is necessary to reproduce the bug in our side and identify quickly the failure unless the bug occurs independently of the input data (in the latter case, please do not us your data but do indicate that the bug is not specific to any data). These data will not be shared, distributed, or duplicated without your expressed and written permission. These data will be destroyed as soon as the bug fix is completed and released. You may also upload your data in a platorm (such as Dropbox) and share privately with us the link to access them;
3) All parameters or the full command line used when the bug occurred (including the variant, CLARK/CLARK-l/CLARK-S);
4) The platform information on which you run CLARK and the version of CLARK ("uname --all" command).
If your bug/error report fails to detail/include all of these information then we do not guarantee a rapid analysis and response to your issue. Please remember that CLARK is being supported so make sure your report is as complete/detailed as possible.