Recurrent neural network-based variant pathogenicity classifier
The technology disclosed relates to constructing a convolutional neural network-based classifier for variant classification. In particular, it relates to training a convolutional neural network-based classifier on training data using a backpropagation-based gradient update technique that progressively match outputs of the convolutional neural network-based classifier with corresponding ground truth labels. The convolutional neural network-based classifier comprises groups of residual blocks, each group of residual blocks is parameterized by a number of convolution filters in the residual blocks, a convolution window size of the residual blocks, and an atrous convolution rate of the residual blocks, the size of convolution window varies between groups of residual blocks, the atrous convolution rate varies between groups of residual blocks. The training data includes benign training examples and pathogenic training examples of translated sequence pairs generated from benign variants and pathogenic variants.
1. A variant pathogenicity classifier, comprising:
a recurrent neural network which runs on numerous processors coupled to memory and is trained to determine pathogenicity of variants;
wherein the recurrent neural network is trained using, as input, benign training example pairs and pathogenic training example pairs of reference protein sequences and alternative protein sequences;
wherein the alternative protein sequences are generated from benign variants and pathogenic variants;
wherein the benign variants include common human missense variants and non-human primate missense variants occurring on alternative non-human primate codon sequences that share matching reference codon sequences with humans; and
wherein the recurrent neural network is trained to capture long-range dependencies in the reference protein sequences and in the alternative protein sequences.
2. The variant pathogenicity classifier of claim 1 , wherein the recurrent neural network is a long short-term memory (LSTM) network.
3. The variant pathogenicity classifier of claim 2 , wherein the recurrent neural network is a gated recurrent unit (GRU) network.
4. The variant pathogenicity classifier of claim 1 , wherein the recurrent neural network is a multi-dimensional recurrent neural network.
5. A variant pathogenicity classifier, comprising:
a recurrent neural network which runs on numerous processors coupled to memory and is trained to determine pathogenicity of variants;
wherein the recurrent neural network is trained using, as input, benign training example pairs and pathogenic training example pairs of reference protein sequences and alternative protein sequences;
wherein the alternative protein sequences are generated from benign variants and pathogenic variants; and
wherein the benign variants include common human missense variants and non-human primate missense variants occurring on alternative non-human primate codon sequences that share matching reference codon sequences with humans.
6. The variant pathogenicity classifier of claim 5 , wherein the recurrent neural network is a long short-term memory (LSTM) network.
7. The variant pathogenicity classifier of claim 5 , wherein the recurrent neural network is a gated recurrent unit (GRU) network.
8. The variant pathogenicity classifier of claim 5 , wherein the recurrent neural network is a multi-dimensional recurrent neural network.
9. A system, comprising:
a variant pathogenicity classifier which runs on numerous processors coupled to memory and is trained to determine pathogenicity of variants;
wherein the variant pathogenicity classifier is trained using, as input, benign training example pairs and pathogenic training example pairs of reference protein sequences and alternative protein sequences;
wherein the alternative protein sequences are generated from benign variants and pathogenic variants; and
wherein the benign variants include common human missense variants and non-human primate missense variants occurring on alternative non-human primate codon sequences that share matching reference codon sequences with humans.
10. The system of claim 9 , wherein the variant pathogenicity classifier is a recurrent neural network.
11. The system of claim 10 , wherein the variant pathogenicity classifier captures long-range dependencies in the reference protein sequences and in the alternative protein sequences.
12. The system of claim 11 , wherein the variant pathogenicity classifier is a long short-term memory (LSTM) network.
13. The system of claim 11 , Wherein the variant pathogenicity classifier is a gated recurrent unit (GRU) network.
14. The system of claim 11 , wherein the variant pathogenicity classifier is a multi-dimensional recurrent neural network.
15. The system of claim 9 , Wherein the variant pathogenicity classifier is a deep spatio-temporal neural network.
16. The system of claim 9 , wherein the variant pathogenicity classifier uses convolutional auto-encoders.