(11 intermediate revisions not shown) | |||
Line 4: | Line 4: | ||
! Code | ! Code | ||
! Level | ! Level | ||
+ | ! Evidence | ||
! align="left" | Meaning | ! align="left" | Meaning | ||
- | |- align="center" | + | |- align="center" valign="top" |
| black | | black | ||
| 1 | | 1 | ||
+ | | none | ||
| align="left" | no observation of a protein has reached a set scoring threshold | | align="left" | no observation of a protein has reached a set scoring threshold | ||
|- align="center" | |- align="center" | ||
| red | | red | ||
| 2 | | 2 | ||
+ | | poor | ||
| align="left" | at least one observation of a protein has exceeded the scoring threshold | | align="left" | at least one observation of a protein has exceeded the scoring threshold | ||
- | |- align="center" | + | |- align="center" valign="top" |
| align="left" | yellow | | align="left" | yellow | ||
| 3 | | 3 | ||
- | | | + | | modest |
- | |- align="center" | + | | align="left" | multiple observations of a protein have a set of common peptides and the distribution of scores for at least one peptide in that set exceeds a minimum test for non-randomness |
+ | |- align="center" valign="top" | ||
| align="left" | green | | align="left" | green | ||
| 4 | | 4 | ||
+ | | good | ||
| align="left" | the set of common peptides contains at least one peptide with a scoring distribution that exceeds a stricter test for non-randomness | | align="left" | the set of common peptides contains at least one peptide with a scoring distribution that exceeds a stricter test for non-randomness | ||
|} | |} | ||
+ | |||
+ | The method for calculating these evidence codes in Perl can be found in the subroutine "gpmdbProteinEvidence" in [ftp://ftp.thegpm.org/repos/gpmdb_rest/gpmdb_rest.pl gpmdb_rest.pl]. | ||
+ | |||
+ | The evidence code associated with any protein accession number can be retrieved using the GPMDB REST interface with a URL call, for example, the evidence code for the mouse protein ENSMUSP00000026459 can be obtained at: | ||
+ | |||
+ | [http://rest.thegpm.org/1/protein/evidence/acc=ENSMUSP00000026459 http://rest.thegpm.org/1/protein/evidence/acc=ENSMUSP00000026459] | ||
+ | |||
+ | The current version of the algorithm for determining these evidence codes for a protein (NBS v.2) is summarized below. | ||
+ | |||
+ | # Retrieve all of the observed E-values for peptides assigned to the protein of interest; | ||
+ | # Generate a frequency histogram for each of the peptides, using bins based on the observations' log(E); | ||
+ | #* all appropriate observations are binned using the values ⌈log(E)⌉ = -2, -3, -4, ... -14 (for example, [http://gpmdb.thegpm.org/protein/model/ENSP00000339186 GRAP2]); | ||
+ | # Any peptide that has not been observed with an E-value < 0.01, report EC = 1; | ||
+ | # Calculate the weighted mean, skew and excess kurtosis for each of the resulting histograms; | ||
+ | # For each peptide that has been observed ≥ 5 times: | ||
+ | ## If the skew and excess kurtosis for the peptide are both ≤ 1.5 or the weighted mean ≤ -5.5, report EC = 4; | ||
+ | ## If either the skew or excess kurtosis for the peptide is ≤ 1.5 or the weighted mean ≤ -3.5, report EC = 3; | ||
+ | # For all other peptides, report EC = 2; | ||
+ | # Report the highest peptide EC as the protein EC. |
GPMDB uses an evidence code system to rate the current observation status of individual protein sequences. The same system is used for gene loci, where applicable. The ratings are as follows:
Code | Level | Evidence | Meaning |
---|---|---|---|
black | 1 | none | no observation of a protein has reached a set scoring threshold |
red | 2 | poor | at least one observation of a protein has exceeded the scoring threshold |
yellow | 3 | modest | multiple observations of a protein have a set of common peptides and the distribution of scores for at least one peptide in that set exceeds a minimum test for non-randomness |
green | 4 | good | the set of common peptides contains at least one peptide with a scoring distribution that exceeds a stricter test for non-randomness |
The method for calculating these evidence codes in Perl can be found in the subroutine "gpmdbProteinEvidence" in gpmdb_rest.pl.
The evidence code associated with any protein accession number can be retrieved using the GPMDB REST interface with a URL call, for example, the evidence code for the mouse protein ENSMUSP00000026459 can be obtained at:
http://rest.thegpm.org/1/protein/evidence/acc=ENSMUSP00000026459
The current version of the algorithm for determining these evidence codes for a protein (NBS v.2) is summarized below.