Protein Info for QEN71_RS13075 in Paraburkholderia sabiae LMG 24235

Annotation: methyl-accepting chemotaxis protein

These analyses and tools can help you predict a protein's function, but be skeptical. For enzymes, over 10% of annotations from KEGG or SEED are probably incorrect. For other types of proteins, the error rates may be much higher. MetaCyc and Swiss-Prot have low error rates, but the best hits in these databases are often quite distant, so this protein's function may not be the same. TIGRFam has low error rates. Finally, many experimentally-characterized proteins are not in any of these databases. To find relevant papers, use PaperBLAST.

Protein Families and Features

1 50 100 150 200 246 transmembrane" amino acids 6 to 25 (20 residues), see Phobius details PF00015: MCPsignal" amino acids 152 to 238 (87 residues), 59.1 bits, see alignment E=2.7e-20

Best Hits

KEGG orthology group: None (inferred from 91% identity to bph:Bphy_3589)

Predicted SEED Role

"Signal transduction histidine kinase CheA (EC 2.7.3.-)" in subsystem Bacterial Chemotaxis or Flagellar motility or Two-component regulatory systems in Campylobacter (EC 2.7.3.-)

Isozymes

Compare fitness of predicted isozymes for: 2.7.3.-

Use Curated BLAST to search for 2.7.3.-

Sequence Analysis Tools

PaperBLAST (search for papers about homologs of this protein)

Search CDD (the Conserved Domains Database, which includes COG and superfam)

Search structures

Predict protein localization: PSORTb (Gram-negative bacteria)

Predict transmembrane helices and signal peptides: Phobius

Check the current SEED with FIGfam search

Find homologs in fast.genomics or the ENIGMA genome browser

Find the best match in UniProt

Protein Sequence (246 amino acids)

>QEN71_RS13075 methyl-accepting chemotaxis protein (Paraburkholderia sabiae LMG 24235)
MVDQYVIAAAVAGSAVGALAAVFAHRWWYGRKAVAQQATALDSQNKALEQAEERHAAEIR
QHEQQARELHALVETLRDDVAEQSVSAEDARAALNALQAQKDEWLQQATRLATEAGRLRH
LSATFERWHEQMISLMAQNHDMHSKNNELSSIVRHVLIVSLNASIEAARAGAAGRGFSIV
ASEVRTLAARSQELSKSYHDSLNRNDLITTATFQDIQAGGKMIAASLASVESLAGQIQSR
LEQATT