Most proteins are composed of one or more conserved functional regions called
domains. This track shows the high-quality, manually-curated
domains found in transcripts located in the GENCODE Genes track by the software HMMER3.
Display Conventions and Configuration
This track follows the display conventions for
The sequences from the knownGenePep table (see
GENCODE Genes description page)
are submitted to the set of Pfam-A HMMs which annotate regions within the
predicted peptide that are recognizable as Pfam protein domains. These regions
are then mapped to the transcripts themselves using the
pslMap utility. A complete shell script log for every version of UCSC genes can be found in
our GitHub repository under
mm10.knownGenes17.csh is for the database mm10 and version 17 of UCSC known genes.
Of the several options for filtering out false positives, the "Trusted cutoff (TC)"
threshold method is used in this track to determine significance. For more information regarding
thresholds and scores, see the HMMER
results interpretation pages.
Note: There is currently an undocumented but known HMMER problem which results in lessened
sensitivity and possible missed searches for some zinc finger domains. Until a fix is released for
HMMER /PFAM thresholds, please also consult the "UniProt Domains" subtrack of the UniProt
track for more comprehensive zinc finger annotations.
pslMap was written by Mark Diekhans at UCSC.
Finn RD, Mistry J, Tate J, Coggill P, Heger A, Pollington JE, Gavin OL, Gunasekaran P, Ceric G,
Forslund K et al.
The Pfam protein families database.
Nucleic Acids Res. 2010 Jan;38(Database issue):D211-22.
PMID: 19920124; PMC: PMC2808889