The GLIMMER 2 system was developed at The Institute for Genomic Research (TIGR) to find genes in microbial genomes. It creates a model of known (or highly probable) genes from a genome and uses the model to locate unknown genes in that genome.

The model is an extension of interpolated Markov models known as an interpolated context model (ICM). It can be built quickly using the genomic sequence itself as a training set.

Predicting genes using GLIMMER is a multistep process:

• obtain a list of the start and stop coordinates of long open reading frames in the genome (long-orfs)
• use the start/stop coordinates to extract the sequence data for these ORFs into a file (extract)
• use the ORFs as a training set to develop an ICM (build-icm)
• use the ICM to predict the positions of genes in the genome (glimmer2)

To increase the accuracy of GLIMMER's prediction, use the ORFs obtained from extract as query sequences for a database search, and eliminate those that do not show matches with known genes from other organisms. Run build-icm using the remaining confirmed genes.

If you use the GLIMMER system as part of published research, please reference the following paper:

• A.L. Delcher, D. Harmon, S. Kasif, O. White, and S.L. Salzberg (1999), Improved microbial gene identification with GLIMMER, Nucleic Acids Research, 27(23): 4636-4641.

Additional references:

• S.L. Salzberg, A.L. Delcher, S. Kasif, and O. White (1998), Microbial gene identification using interpolated Markov models, Nucleic Acids Research, 26(2): 544-548.

The programs of the GLIMMER system are restricted to local access.

• long-orfs: returns a list of the start and stop positions of long open reading frames
• extract: extracts ORF sequences from a genomic sequence based on start/stop coordinates
• build-icm: creates a genome-specific ICM from a training set of ORFs
• glimmer2: finds genes in microbial genomic sequences based on an ICM created for that genome

This website will look much better in a browser that supports web standards, but it has been designed so that it is still usable and accessible to any browser or web-enabled device.