Automated eukaryotic gene structure annotation using EVidenceModeler and the Program to Assemble Spliced Alignments

Brian J Haas¹^,2 , Steven L Salzberg³ , Wei Zhu¹ , Mihaela Pertea³ , Jonathan E Allen³^,4 , Joshua Orvis¹^,5 , Owen White¹^,5 , C Robin Buell¹^,6 and Jennifer R Wortman¹^,5

¹J Craig Venter Institute, The Institute for Genomic Research, Rockville, 9712 Medical Center Drive, Maryland 20850, USA

²Broad Institute of MIT and Harvard, 7 Cambridge Center, Cambridge, Massachusetts 02142, USA

³Center for Bioinformatics and Computational Biology, Department of Computer Science, 3125 Biomolecular Sciences Bldg #296, University of Maryland, College Park, Maryland 20742, USA

⁴Computation Directorate, Lawrence Livermore National Laboratory, 7000 East Avenue, Livermore, California 94550, USA

⁵Institute for Genome Sciences, University of Maryland Medical School, Baltimore, Maryland 21201, USA

⁶Department of Plant Biology, Michigan State University, East Lansing, Michigan 48824, USA

author email corresponding author email

Genome Biology 2008, 9:R7doi:10.1186/gb-2008-9-1-r7


Published:	11 January 2008

Subject areas: Bioinformatics, Genome studies, Methods

Additional files

Additional file 1:

Supplementary figures. (Figure S1 shows the difference in rice gene prediction accuracy between using trained and intuitively set evidence weights. Figure S2 shows the change in human gene prediction accuracy due to application of PASA. Figure S3 shows the comparison of 1,058 reference gene structure exon distribution to all rice gene annotations. Figure S4 shows the gene prediction accuracies for EGASP gene sets. Figure S5 shows filtering EVM predictions with low support. Figure S6 shows optimization of evidence weights by exploring weight and evidence combinations.

Format: PDF Size: 314KB Download file

This file can be viewed with: Adobe Acrobat Reader

Additional file 2:

Supplementary data tables. Table S1 provides trained weights for evidence based on evaluating 500 rice gene structures. Table S2 shows the gene prediction accuracy for EVM measured using 500 reference rice gene structures. Table S3 provides trained EVM weights including PASA. Table S4 provides trained EVM evidence weights for the ENCODE regions. Table S5 shows the EVM prediction accuracy using trained evidence weights for ENCODE regions.

Format: PDF Size: 155KB Download file

This file can be viewed with: Adobe Acrobat Reader