Deep learning-based techniques for pre-training deep convolutional neural networks

A system including one or more processors coupled to memory, the memory loaded with computer instructions to reduce overfitting of a neural network-implemented model that processes sequences of amino acids and accompanying position frequency matrices (PFMs), the instructions, when executed on the processors, implement actions comprising:

• generating supplemental training example sequence pairs, labelled benign, that include a start location, through a target amino acid location, to an end location, wherein each supplemental training example sequence pair:
• matches the start location and the end location of a missense training example sequence pair;
• and
• has identical amino acids in a reference and an alternate sequence of amino acids;
• inputting with each supplemental training example sequence pair a supplemental training PFM that is identical to the PFM of the missense training example sequence pair at the matching start and end location;
• and
• training the neural network-implemented model using the benign supplemental training example sequence pairs, the supplemental training PFMs, the missense training example sequence pairs, and the PFMs of the missense training example sequence pairs at the matching start and end locations;
• whereby training influence of the supplemental training PFMs is attenuated or counteracted during the training.

A non-transitory computer readable storage medium impressed with computer program instructions to reduce overfitting of a neural network-implemented model that processes sequences of amino acids and accompanying position frequency matrices (PFMs), the instructions, when executed on a processor, implement a method comprising:

• generating supplemental training example sequence pairs, labelled benign, that include a start location, through a target amino acid location, to an end location, wherein each supplemental training example sequence pair:
• matches the start location and the end location of a missense training example sequence pair;
• and
• has identical amino acids in a reference and an alternate sequence of amino acids;
• inputting with each supplemental training example sequence pair a supplemental training PFM that is identical to the PFM of the missense training example sequence pair at the matching start and end location;
• and
• training the neural network-implemented model using the benign supplemental training example sequence pairs, the supplemental training PFMs, missense training example sequence pairs, and the PFMs of the missense training example sequence pairs at the matching start and end locations;
• whereby training influence of the supplemental training PFMs is attenuated during the training.