What is FragSort?

This downloadable Windows program compares terminal restriction fragment sizes (TRFs) obtained from microbial community DNA amplified with a labeled primer, and digested with one, two, three, or four different restriction enzymes to the TRFs generated from in silico amplification and digestions of a gene sequence database (e.g. 16S rRNA gene database). Four example databases are provided, however custom databases using your primer and restriction enzymes can be easily created using the MICA program.

Input files consist of two, three, or four different text files containing the fragment sizes and normalized peak areas (or heights) from each individual restriction digestion. As an example, consider the following TRFLP data from a Hha1 digestion of environmental DNA amplified with a labeled 8f primer.

There are peaks at 61, 72, 81, 171, 204, 209, 221, 338, 372, 390, 512, 550, 565 and 918 bp. The respective peak areas are 125, 195, 55, 60, 835, 160, 50, 165, 1250, 125, 200, 55, 970, and 755 fluorescence units. This data can be copied directly from Genescan output to a text or comma separated variable (.csv) file. The normalized peak areas are calculated by dividing individual peak areas by the sum of all peak areas (or heights). This would give respectively, 2.5%, 3.9%, 1.1%, 1.2%, 16.7%, 3.2%, 1.0%, 3.3%, 25.0%, 2.5%, 4.0%, 1.1%, 19.4% and 15.1% for the data above. The format of the FRAGSORT input file would be the following:

No extra characters or returns are permitted in the text (or .csv) file. Similar files would be created based on 8f-Rsa1 and 8f-Msp1 TRFLP data. Example files are available for download in the downloads link. A program included in FRAGSORT called NORMALIZER automatically converts a GENESCAN multiple lane output file into FRAGSORT input files.

The FRAGSORT program output is a list of microorganisms and TRF sizes that correlate with multiple experimentally generated TRFLP profiles. The organisms are initially listed in descending order from the greatest to the least normalized TRF peak area (or height). The output can be sorted based on organism name, TRF areas, Accession ID, or minimum TRF area. Fragment sizing error can adjusted by the user to either a constant value or a variable value based on fragment size. The program can be set to output all organisms that match 2 of 3 TRFLP fingerprints, or all three fingerprints. Data can be saved for later analysis and/or copied to other programs. See the features page for more advanced features of this program.