Meta-MEME motif search tools
Meta-MEME provides a suite of programs for searching a sequence database for occurrences of a known collection of motifs. The programs can search DNA or protein databases, and can search for individual motifs, or for ordered series or unordered clusters of motifs.
- fimo: Searches a sequence database for occurrences of known motifs. This program treats each motif independently and reports all putative motif occurrences below a specified p-value threshold.
- mast: Searches a sequence database for occurences of known motifs. This program assumes exactly one occurrence of a motif per sequence, and each sequence in the database is assigned a p-value, which is based on the product of the p-values of the individual motif occurrences in that sequence.
- beadstring: Searches a sequence database for a given collection of motifs. The motifs are represented using a hidden Markov model with a linear topology, thus requiring that the motifs appear in a specified order.
- miao: Searches a sequence database for clusters of known motifs. As in
beadstring
motifs are represented using a hidden Markov model, butmiao
uses the complete or star topologies. This allows the motifs to appear in any order.- mcast: Searches a sequence database for clusters of known motifs. Like beadstring, and miao,
mcast
employs a motif-based hidden Markov model, but uses a star topology and a novel scoring algorithm. The motifs may appear in any order.Inputs:
Each program takes as input a collection of DNA or protein motifs in an XML format, and a databse of sequences in FASTA format.Outputs:
An XML file using the CisML schema.
General options (available in all programs)
Program specific optionsA number of options are available for each of the programs. Some of the options may be used with any of the programs:
--bg-file <bfile>
- Read background frequencies from <bfile>. The default is to calculate the frequencies from the database. If the argument is the keywordmotif-file
the frequencies will be taken from the motif file.--max-seqs <max>
- Print results for no more than <max> sequences (default: all).--motif <id>
- Use only the motif identified by <id>. This option may be repeated.--progress <value>
- Print to standard error a progress message after every <value> seconds.--verbosity 1|2|3|4
- Set the verbosity of status reports to standard error. The default level is 2.Other options are specific to the individual program. This table describes which options are available for which programs.
Other Tools
The Meta-MEME software distribution also includes the following other programs:
li> ceqlogo: Creates logos in EPS or SVG format from motifs.
- tomtom: Measure the similarity of motifs.
- mhmm: Build a motif based HMM.
- mhmm2html: Convert the text output from one of the five primary programs into HTML format.
- mhmme: Randomly generate sequences according to a given Meta-MEME model.
- draw-mhmm: Convert a motif-based HMM into a format suitable for drawing by the graphviz program from AT&T Research.
- fasta-get-markov: Estimate a Markov model from a FASTA file of sequences.
- qvalue: Compute q-values from p-values.
- transfac2meme: Converts a TRANSFAC matrix file to MEME output format.
- clustalw2fasta: Converts a Clustalw multiple alignment into FASTA format.
Additional documentation is available concerning
- the format for the FASTA sequence files required by mhmms and mhmmscan(with a sample),
- the format for the MEME file required by mhmm (with a sample),
- the format for the background file,
- the format for the TRANSFAC file used by mcast and transfac2meme, and
- how Meta-MEME determines motif order and spacing information from MEME output.
Sample output files are available for
- the motif-based HMMs produced by mhmm, and
- the database search results produced by mhmms.
Visit the Meta-MEME home page at the San Diego Supercomputer Center.
The programs are:
Meta-MEME was developed by William Stafford Noble in the Department of Genome Sciences the University of Washington, and by Timothy Bailey at the University of Queensland, with input from Charles Elkan in the Department of Computer Science and Engineering at the University of California, San Diego and Michael Gribskov at the San Diego Supercomputer Center. Meta-MEME is funded by the National Biomedical Computation Resource.
Copyright information. Please send comments and questions to Charles Grant at cegrant@u.washington.edu