Prediction of protein function from sequence analysis Rita Casadio BIOCOMPUTING GROUP University of Bologna, Italy

The “omic” era Genome Sequencing Projects: Archaea : 74 species In Progress:52 Bacteria: 973 species In Progress: 2266 species Complete-23 Draft Assembly–318 In Progress-359 Eukaryotic: http://www.ncbi.nlm.nih.gov/genomes/static/gpstat.html Update: January 2010

The Data Bases of Biological Sequences and Structures GenBank: 108,431,692 sequences 106,533,156,756 nucleotides NR(*): 10,381,779 sequences 3,542,056,219 residues >BGAL_SULSO BETA-GALACTOSIDASE Sulfolobus solfataricus. MYSFPNSFRFGWSQAGFQSEMGTPGSEDPNTDWYKWVHDPENMAAGLVSG DLPENGPGYWGNYKTFHDNAQKMGLKIARLNVEWSRIFPNPLPRPQNFDE SKQDVTEVEINENELKRLDEYANKDALNHYREIFKDLKSRGLYFILNMYH WPLPLWLHDPIRVRRGDFTGPSGWLSTRTVYEFARFSAYIAWKFDDLVDE YSTMNEPNVVGGLGYVGVKSGFPPGYLSFELSRRHMYNIIQAHARAYDGI KSVSKKPVGIIYANSSFQPLTDKDMEAVEMAENDNRWWFFDAIIRGEITR GNEKIVRDDLKGRLDWIGVNYYTRTVVKRTEKGYVSLGGYGHGCERNSVS LAGLPTSDFGWEFFPEGLYDVLTKYWNRYHLYMYVTENGIADDADYQRPY YLVSHVYQVHRAINSGADVRGYLHWSLADNYEWASGFSMRFGLLKVDYNT KRLYWRPSALVYREIATNGAITDEIEHLNSVPPVKPLRH 35,5 HGE! SwissProt: 514,212 sequences 180,900,945 residues PDB: 60,654 structures membrane proteins <2% Update: January 2009 (*) CDS translations+PDB+SwissProt+PIR+PRF

From Genotype to Phenotype …with different effects depending on variability Over 20 millions of single mutations are known in genes >protein kinase acctgttgatggcgacagggactgtatgctgatctatgctgatgcatgcatgctgactactgatgtgggggctattgacttgatgtctatc.... Genes in DNA... (about 30,000 in the human genome) …code for proteins... ….in methabolic pathways From Genotype to Phenotype …proteins correspond to functions... …when they are expressed From 5000 to 10000 proteins per tissue Proteins interact

The Human Interactome in STRING STRING 8—a global view on proteins and their functional interactions in 630 organisms- Jensen et al., 2009, Nucleic Acids Research, Vol 37. The Human Interactome in STRING 22,937 proteins and 1,482,533 interactions http://string.embl.de

One problem of the “omic era”: Protein functional annotation

The Protein Data Bank No of Proteins with known structure: 57529 http://www.rcsb.org/pdb/home/home.do No of Proteins with known structure: 57529

SCOP: Structural Classification of Proteins Domains are hierarchically classified: class - fold: proteins with secondary structures in same arrangement with the same topological connections superfamily: structures and functional features suggest a common evolutionary origin family: proteins with identities ≥30%; with identities <30% but with similar structures and functions

From the Protein Sequence to the Structure and Function space Lesk A., 2004

From the Protein Sequence to the Structure space PDB Sequence comparison Sequence Identity (%) 0% 30% 100% Fold recognition Machine-learning aided alignment Threading New Folds Ab initio and de novo modelling Machine-learning prediction of structural features From the Protein Sequence to the Structure space

What is protein function? From the Protein Sequence to the Structure and Function space What is protein function?

What is a function? For enzymes: function can be defined on the basis of the catalysed molecular reaction. e.g. aspartic aminotransferase (AST)

In biochemistry, a transaminase or an aminotransferase is an enzyme that catalyzes a type of reaction between an amino acid and an α-keto acid. Specifically, this reaction (transamination) involves removing the amino group from the amino acid, leaving behind an α-keto acid, and transferring it to the reactant α-keto acid and converting it into an amino acid. The enzymes are important in the production of various amino acids, and measuring the concentrations of various transaminases in the blood is important in the diagnosing and tracking many diseases. Transaminases require the coenzyme pyridoxal-phosphate, which is converted into pyridoxamine in the first phase of the reaction, when an amino acid is converted into a keto acid. Enzyme-bound pyridoxamine in turn reacts with pyruvate, oxaloacetate, or alpha-ketoglutarate, giving alanine, aspartic acid, or glutamic acid, respectively. The presence of elevated transaminases can be an indicator of liver damage.

Enzyme Commission (E.C.) classification A hierarchical classification for enzymes

EC 2.6 Transferring nitrogenous groups EC 2.6.1Transaminases EC 2.6.1.1 Aspartate transaminase Other name(s): glutamic-oxaloacetic transaminase; glutamic-aspartic transaminase; transaminase A; AAT; AspT; 2-oxoglutarate-glutamate aminotransferase; aspartate α-ketoglutarate transaminase; aspartate aminotransferase; aspartate-2-oxoglutarate transaminase; aspartic acid aminotransferase; aspartic aminotransferase; aspartyl aminotransferase; AST; glutamate-oxalacetate aminotransferase; glutamate-oxalate transaminase; glutamic-aspartic aminotransferase; glutamic-oxalacetic transaminase; glutamic oxalic transaminase; GOT (enzyme); L-aspartate transaminase; L-aspartate-α-ketoglutarate transaminase; L-aspartate-2-ketoglutarate aminotransferase; L-aspartate-2-oxoglutarate aminotransferase; L-aspartate-2-oxoglutarate-transaminase; L-aspartic aminotransferase; oxaloacetate-aspartate aminotransferase; oxaloacetate transferase; aspartate:2-oxoglutarate aminotransferase; glutamate oxaloacetate transaminase Systematic name: L-aspartate:2-oxoglutarate aminotransferase

Problems: Isoforms e.g How to differentiate the function of the cytoplasmic aspartate amintransferase from that of mitochondrial isoform? Non enzymatic proteins

GO function vocabulary: The Ontologies Cellular component Biological process Molecular function GO function vocabulary: http://www.geneontology.org/

Gene Ontology classification: The human cytoplasmic aspartate transaminase GO:0004069 GO:0005829 GO:0006533

One BIG problem of the “omic era”: Protein functional annotation

Functional annotation in silico by homology search ADH1_SULSO ----------MRAVRLVEIGKP--LSLQEIGVPKPKGPQVLIKVEAAGVCHSDVHMRQGRFGNLRIVE ADH_CLOBE ----------MKGFAMLGINKLG---WIEKERPVAGSYDAIVRPLAVSPCTSDIHTVFEGA------- ADH_THEBR ----------MKGFAMLSIGKVG---WIEKEKPAPGPFDAIVRPLAVAPCTSDIHTVFEGA------- ADH1_SOLTU MSTTVGQVIRCKAAVAWEAGKP--LVMEEVDVAPPQKMEVRLKILYTSLCHTDVYFWEAKG------- ADH2_LYCES MSTTVGQVIRCKAAVAWEAGKP--LVMEEVDVAPPQKMEVRLKILYTSLCHTDVYFWEAKG------- ADH1_ASPFL ----MSIPEMQWAQVAEQKGGP--LIYKQIPVPKPGPDEILVKVRYSGVCHTDLHALKGDW------- Sequence comparison is performed with alignment programs Sequence identity  40 % Similar structure and function (??) Methods for similarity searches: BLAST, Psi-BLAST (http://www.ncbi.nlm.nih.gov/BLAST/) sequence Altschul et al., (1990) J Mol Biol 215:403-410 Altschul et al., (1998) Nucleic Acids Res. 25:3389-3402 Pfam (http://pfam.wustl.edu/hmmsearch.shtml) sequence/structure Bateman et al., (2000) Nucleic Acids Research 28:263-266

Transfer by inheritance: Function annotation transfer from sequence through homology

http://www.uniprot.org/

PDB The annotation process at UniProt

Open problems of “inheritance through homology “ Not all UniProt files are GO annotated The optimal threshold value of sequence identity for function transfer is not known Proteins contain multiple domains Proteins can share common domains and not necessarily the same function In proteins different combination of shared domains lead to different biological roles