MHC & Prediction Algorithm

(Threshold Vs Best Scoring Subsequence Distribution Profile)

An example out put of Antigen 85B (M. tuberculosis).

The figure appears only during sub sequence analysis. Here the highest scoring peptide frame is treated as representative of whole query sub sequence. Top red line is the entire threshold range and the blue line against each allele represents threshold value at or above which the query sub sequence will be predicted as binder.
The figure is useful for experienced users in selecting threshold, while locating promiscuous binding regions.


Sequence :          FFVASDREPLMN
Sub sequences: FFVASDREP                   Score: 1.2
                                  FVASDREPL                 Score: 1.0
                                     VASDREPLM             Score: 1.4
                                        ASDREPLMN          Score: 0.8

The highest scorer i.e. " VASDREPLM " will be plotted as representative of sequence. The blue line represents the threshold at or below which this subsequent will be predicted as binder.


But to me the other two figures are more informative because
i) They are not restricted to highest scoring peptide frame hence will deliver true picture of subsequent.
ii) The location and score of highest scoring subsequent can be determined.
iii) Other binding frames close to threshold or generated by one or two amino acid shift in highest scoring sequence can also be easily located.
MHC & Prediction Algorithm