Skip to main content

Table 5 Transfer learning methods for representative MHC alleles

From: In silico design of MHC class I high binding affinity peptides through motifs activation map

MHC allel

SRCC

AUC

high affinity rate1

transfer learning method2

Improvement3

A*6802

0.499

0.817

0.9

fine-tune

0.61

 

0.058

0.537

0.29

zero-shot

 

A*0206

0.458

0.778

0.75

fine-tune

0.08

 

0.378

0.73

0.67

zero-shot

 

B*2705

0.701

0.929

0.92

fine-tune

0.92

 

0.167

0.602

0

zero-shot

 

Mamu-A1*001:01

0.755

0.943

0.839

fine-tune

0.809

 

0.256

0.65

0.03

zero-shot

 

A*0201

0.117

0.576

0.859

originally

 
  1. *The length of the peptides for training transfer learning is nine.
  2. 1The high-affinity is scored by http://tools.iedb.org/mhci/.
  3. 2There are two types of transfer learning. Fine-tune indicates keeping training on the basis of the original model with a smaller learning rate (\(\frac {1}{10}\) learning rate). Zero-shot indicates that direct transferring without further training. Originally indicates the model is the well-trained model that all the transfer learning is based on.
  4. 3the Improvement stands for the increase of high-affinity rate from zero-shot transfer method to fine-tune method