Skip to main content

Table 1 Binding data statistics

From: Dataset size and composition impact the reliability of performance benchmarks for peptide-MHC binding predictions

 

BD2009+

BD2013

Blind++

Alleles

79

114

53

Data sets

170

257

90

Data set size

   

Average

792

685

324

Min

50

50

50

Max

6,961

8,826

1,865

Total data points

134,645

176,161

29,169

  1. +All cross-validations were carried out using BD2009.
  2. ++Blind was generated by subtracting BD2009 from BD2013. Against Blind, all blind predictions were made using the predictors trained on BD2009.
  3. Each (MHC, length) combination is associated with a data set.