fjdf / slamem

Finding Maximal Exact Matches (MEMs) using a Sampled LCP Array

slaMEM

slaMEM is a tool used to efficiently retrieve MEMs (Maximal Exact Matches) between a reference genome sequence and one or more query sequences, similarly to these software tools:

• E-MEM
• MUMmer
• essaMEM
• sparseMEM
• backwardMEM

slaMEM relies on an FM-Index together with a new data structure called SSILCP (Sampled Search Intervals from Longest Common Prefixes) to store information about parent intervals in a time- and space-efficient way.

slaMEM also includes an useful feature to display the locations of the found MEMs, generating images like the one below.

Reference

If you use slaMEM, please cite:

Fernandes, Francisco and Ana T. Freitas. slaMEM: efficient retrieval of maximal exact matches using a sampled LCP array. Bioinformatics 30.4 (2014): 464-471.

Manual

Install

make

Usage

./slaMEM (<options>) <reference_file> <query_file(s)>

Options:

• mem : find MEMs: any number of occurrences in both ref and query (default)
• mam : find MAMs: unique in ref but any number in query
• l : minimum match length (default=20)
• o : output file name (default="*-mems.txt")
• b : process both forward and reverse strands
• n : discard 'N' characters in the sequences
• m : minimum sequence size (e.g. to ignore small scaffolds)
• r : load only the reference(s) whose name(s) contain(s) this string

Extra:

• v : generate MEMs map image from this MEMs file

Example:

./slaMEM -b -l 10 ./ref.fna ./query.fna
./slaMEM -v ./ref-mems.txt ./ref.fna ./query.fna