Support Vector Machine GPCR Subfamily Classification Results for Q28659

>Q28659|Q28659 MLATTSARVS SSSLYSFWQP HMVLNGAARA PQSWVPQMNG LRLGLVPGLS RVRLGTMLLW GMIFLPSIYM ELVHYSSYEM VIPESLTVEG SEKPGEKASY ILFMQGQKQL VHLHVKRDYF VDDFPVYSYH NRVLGQEMLF ISRNCYYEGY IDGVPGSFVS VNTCSGLRGV LVKGETSYSI EPILSSKRFE HALYTMAHGA HVSCSVTSKG GQGMSTSRQQ GSRKLHNPQA LSYLWSHTKS VEMFVVVDNQ RFQMWGRNVS ETVQRVMDII ALANSFTRGI NTEVVLAGME IWTEGDLTEV AADLQVTLRN FNSWRQEQLV HRVRHDVAHM IVGRHPGENT GQAFLNGACS SGFAAAVESF HHEDILLFAA LMAHELGHNL GIQHDHSACT CKNQPFCLMG ENITKESSFS NCSSDDFYRF LREHRGACLF NKPRHRSRTR RLSRCGNGVV ETPEQCDCGS DCALDPCCDS MCRLKDNAQC GYGLCCFRCK YRRKGFLCRS IRGNCDLPEY CSGKSASCPP DAYKQDGTPC DRVYRCLGGQ CMNPDKQCSN IYGIPARSAP EECYVLMNSK GDRFGNCGSP PALQSSYVPC ADENIFCGKL ICTEVKLLPQ ILPQHTVIQV AYEDDWCWSI DSNSGCSDYG DVQRNTYCAL NKVCKDHSCV VYQAPNSDCQ ADEMCSGKGV CNNFRHCHCD SGYAPPDCRN PGTGGSVDSG PPGKPYNRNI SSSTNSSRIT KKKKSENLNV LFFVVPIFLI IVLCVLIILS YLWSEVKSVV VSIAESKEES EESSEELPSE ESVEAPPPEQ PAQQQEAPQQ QEAPPAREAP PPEAARPAEA PPPPEQAPPP EQAPPPEAPK PAEAPPPPEA APPPQAPPPP EAAPPPEAPP PPEAAPPPEA PPPPEAAPPP EAPPPPEAPP PPEAPPPPQ


Here is a list of Class A Rhodopsin like subfamilies with support vector machine models and the score this sequence received with respect to each model.

-0.9519809Gonadotropin-releasing hormone
-1.0128574Nucleotide-like
-1.0695655Hormone protein
-1.1591787Thyrotropin-releasing hormone & Secretagogue
-1.1957017Viral
-1.2328777Melatonin
-1.2532666Prostanoid
-1.283769Cannabis
-1.3350959Olfactory
-1.3580929Platelet activating factor
-1.4001245Lysosphingolipid & LPA (EDG)
-1.4400072(Rhod)opsin
-1.7414204Peptide
-1.752652Amine
-2.0311658Class A Orphan/other
 -1.0102279Mas proto-oncogene
 -1.199689SREB
 -1.2183043Chemokine receptor-like 2
 -1.2198513GPR37 like (peptide receptor)
 -1.2713995ORPH
 -1.2954063GP40 like
 -1.32108Adrenomedullin (G10D)
 -1.327878RDC1
 -1.3392701GPR
 -1.4095948BONZO
 -1.6380992yet even more Class A Orphans


HOW TO INTERPRET THE SCORES:

Subfamily Scoring:

This sequence does not appear to be in any of the subfamilies for which we have models



The support vector machine for each subfamily is trained to score positive examples (members of the subfamily) as 1.0 and negative examples (non-members) as -1.0

If this sequence receives a negative score with respect to a subfamily, it is probably not a member of the subfamily.

If this sequence receives a positive score with respect to a subfamily, it is probably in the subfamily.

In general, the score's distance from zero gives you the confidence level of the prediction. Positive scores greater than +1.0 mean a classifier has strongly accepted your sequence. Negative scores less than -1.0 mean a classifier has strongly rejected your sequence. For a more detailed evaluation of scores and confidence levels, take a look at Classification Statistics.
SOME ADDITIONAL INFORMATION:

There were 17 Class B Secretin like and 11 Class C Metabotropic glutamate / pheromone and 2 Class D Fungal pheromone and 5 Frizzled/Smoothened family and 67 Class A Rhodopsin like subfamilies whose classifiers were not run, because the families or subfamilies that contain them were rejected by classifiers higher in the hierarchy. For example, if the Amine receptor classifier scores this sequence negatively, it will not be checked with respect to the subfamilies of Amine receptors (Histamine, Serotonin, Dopamine, Octopamine and Acetylcholine (muscarinic) receptors).

Not all subfamilies have models yet. This sequence may be in a subfamily that has not yet been modeled.

Please cite: R. Karchin, K. Karplus and D. Haussler "Classifying G-Protein Coupled Receptors with Support Vector Machines" http://www.cse.ucsc.edu/research/compbio/gpcr-subclass [postscript, pdf]