Limits...
PlantTFcat: an online plant transcription factor and transcriptional regulator categorization and analysis tool.

Dai X, Sinharoy S, Udvardi M, Zhao PX - BMC Bioinformatics (2013)

Bottom Line: These prediction logics effectively distinguish TF/TR/CR families with common conserved domains.Our systematic performance evaluations indicate that PlantTFcat annotates known TF/TR/CR families with high coverage and sensitivity.PlantTFcat provides an analysis tool to identify and categorize plant TF/TR/CR genes on a genomic scale.

View Article: PubMed Central - HTML - PubMed

Affiliation: Plant Biology Division, The Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway 73401 Ardmore, OK, USA. pzhao@noble.org.

ABSTRACT

Background: Plants regulate intrinsic gene expression through transcription factors (TFs), transcriptional regulators (TRs), chromatin regulators (CRs), and the basal transcription machinery. An understanding of plant gene regulatory mechanisms at a systems level requires the identification of these regulatory elements on a genomic scale.

Results: Here, we present PlantTFcat, a high-performance web-based analysis tool that is designed to identify and categorize plant TF/TR/CR genes from genome-scale protein and nucleic acid sequences by systematically analyzing InterProScan domain patterns in protein sequences. The comprehensive prediction logics that are included in PlantTFcat are based on relationships between gene families and conserved domains from 108 published plant TF/TR/CR families. These prediction logics effectively distinguish TF/TR/CR families with common conserved domains. Our systematic performance evaluations indicate that PlantTFcat annotates known TF/TR/CR families with high coverage and sensitivity.

Conclusions: PlantTFcat provides an analysis tool to identify and categorize plant TF/TR/CR genes on a genomic scale. PlantTFcat is freely available to the public at http://plantgrn.noble.org/PlantTFcat/.

Show MeSH
Venn diagrams that show the differences in prediction results between PlantTFcat and BLAST-based methods. PlnTFDB represents the TF/TR dataset downloaded from PlnTFDB, PlantTFcat represents the PlantTFcat predictions, and BLAST represents the BLAST search predictions. (a) The results using the Arabidopsis thaliana gene models release 8 (TAIR8) as the test dataset. (b) The results using the Populus trichocarpa JGI gene models v1.1 as the test dataset.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4225725&req=5

Figure 3: Venn diagrams that show the differences in prediction results between PlantTFcat and BLAST-based methods. PlnTFDB represents the TF/TR dataset downloaded from PlnTFDB, PlantTFcat represents the PlantTFcat predictions, and BLAST represents the BLAST search predictions. (a) The results using the Arabidopsis thaliana gene models release 8 (TAIR8) as the test dataset. (b) The results using the Populus trichocarpa JGI gene models v1.1 as the test dataset.

Mentions: We compared the false positive rate of PlantTFcat with traditional BLAST-based methods. The Arabidopsis thaliana TAIR8 was chosen as a test dataset. A BLAST search (e-value < =1e-04) was run against the TAIR8 TF/TR genes that were downloaded from PlnTFDB. The BLAST search correctly hit all of the 2,757 TF/TR reference genes because the test dataset contains these reference genes. In addition, the BLAST-based method reported 3,870 more homologous genes as TF/TR candidates that had been excluded by PlnTFDB (Figure 3a). In contrast, PlantTFcat reported only 95 false positives, as described above. We also tested these methods against the Populus trichocarpa JGI gene models v1.1 and achieved similar results (Figure 3b). These results suggest that PlantTFcat is a better choice for TF/TR/CR gene annotation over a traditional BLAST search against a reference dataset due to a lower rate of false positives. The details for both comparisons are available in Additional file 4.


PlantTFcat: an online plant transcription factor and transcriptional regulator categorization and analysis tool.

Dai X, Sinharoy S, Udvardi M, Zhao PX - BMC Bioinformatics (2013)

Venn diagrams that show the differences in prediction results between PlantTFcat and BLAST-based methods. PlnTFDB represents the TF/TR dataset downloaded from PlnTFDB, PlantTFcat represents the PlantTFcat predictions, and BLAST represents the BLAST search predictions. (a) The results using the Arabidopsis thaliana gene models release 8 (TAIR8) as the test dataset. (b) The results using the Populus trichocarpa JGI gene models v1.1 as the test dataset.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4225725&req=5

Figure 3: Venn diagrams that show the differences in prediction results between PlantTFcat and BLAST-based methods. PlnTFDB represents the TF/TR dataset downloaded from PlnTFDB, PlantTFcat represents the PlantTFcat predictions, and BLAST represents the BLAST search predictions. (a) The results using the Arabidopsis thaliana gene models release 8 (TAIR8) as the test dataset. (b) The results using the Populus trichocarpa JGI gene models v1.1 as the test dataset.
Mentions: We compared the false positive rate of PlantTFcat with traditional BLAST-based methods. The Arabidopsis thaliana TAIR8 was chosen as a test dataset. A BLAST search (e-value < =1e-04) was run against the TAIR8 TF/TR genes that were downloaded from PlnTFDB. The BLAST search correctly hit all of the 2,757 TF/TR reference genes because the test dataset contains these reference genes. In addition, the BLAST-based method reported 3,870 more homologous genes as TF/TR candidates that had been excluded by PlnTFDB (Figure 3a). In contrast, PlantTFcat reported only 95 false positives, as described above. We also tested these methods against the Populus trichocarpa JGI gene models v1.1 and achieved similar results (Figure 3b). These results suggest that PlantTFcat is a better choice for TF/TR/CR gene annotation over a traditional BLAST search against a reference dataset due to a lower rate of false positives. The details for both comparisons are available in Additional file 4.

Bottom Line: These prediction logics effectively distinguish TF/TR/CR families with common conserved domains.Our systematic performance evaluations indicate that PlantTFcat annotates known TF/TR/CR families with high coverage and sensitivity.PlantTFcat provides an analysis tool to identify and categorize plant TF/TR/CR genes on a genomic scale.

View Article: PubMed Central - HTML - PubMed

Affiliation: Plant Biology Division, The Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway 73401 Ardmore, OK, USA. pzhao@noble.org.

ABSTRACT

Background: Plants regulate intrinsic gene expression through transcription factors (TFs), transcriptional regulators (TRs), chromatin regulators (CRs), and the basal transcription machinery. An understanding of plant gene regulatory mechanisms at a systems level requires the identification of these regulatory elements on a genomic scale.

Results: Here, we present PlantTFcat, a high-performance web-based analysis tool that is designed to identify and categorize plant TF/TR/CR genes from genome-scale protein and nucleic acid sequences by systematically analyzing InterProScan domain patterns in protein sequences. The comprehensive prediction logics that are included in PlantTFcat are based on relationships between gene families and conserved domains from 108 published plant TF/TR/CR families. These prediction logics effectively distinguish TF/TR/CR families with common conserved domains. Our systematic performance evaluations indicate that PlantTFcat annotates known TF/TR/CR families with high coverage and sensitivity.

Conclusions: PlantTFcat provides an analysis tool to identify and categorize plant TF/TR/CR genes on a genomic scale. PlantTFcat is freely available to the public at http://plantgrn.noble.org/PlantTFcat/.

Show MeSH