PANTHER version 14: more genomes, a new PANTHER GO-slim and improvements in enrichment analysis tools

Abstract

PANTHER (Protein Analysis Through Evolutionary Relationships, http://pantherdb.org) is a resource for the evolutionary and functional classification of genes from organisms across the tree of life. We report the improvements we have made to the resource during the past two years. For evolutionary classifications, we have added more prokaryotic and plant genomes to the phylogenetic gene trees, expanding the representation of gene evolution in these lineages. We have refined many protein family boundaries, and have aligned PANTHER with the MEROPS resource for protease and protease inhibitor families. For functional classifications, we have developed an entirely new PANTHER GO-slim, containing over four times as many Gene Ontology terms as our previous GO-slim, as well as curated associations of genes to these terms. Lastly, we have made substantial improvements to the enrichment analysis tools available on the PANTHER website: users can now analyze over 900 different genomes, using updated statistical tests with false discovery rate corrections for multiple testing. The overrepresentation test is also available as a web service, for easy addition to third-party sites.

Figure 1.

Species tree of plant genomes in PANTHER version 14. New plant genomes are in red.

Figure 2.

The updated PANTHER Protein Classes protease and protease inhibitor viewed in PANTHER Prowler. The major subclasses of proteases, such as aspartic protease and metalloprotease, have been aligned with MEROPS, while families in smaller subclasses (e.g. threonine proteases) or proteases of unknown mechanism, remain directly under the upper level classes.

Figure 3.

Phylogenetic distribution of genomes available in PANTHER 14.0.

Figure 4.

User interface to support an additional 800+ genomes in the PANTHER gene list analysis tools. Users can convert their gene list to a UniProt ID list, upload it and select the list type as ‘ID’s from Reference Proteome Genome’. Then select the organism in the drop-down menu.

Figure 5.

PANTHER overrepresentation test results. (A) Analysis options. Users can choose between two different statistical tests, Fisher's Exact and Binomial, and correction methods. (B) A screenshot of a result page using the Fisher's Exact test and FDR correction method. Both the raw P-values and FDR corrected values are reported in the last two columns.