Hi, I was just reading the associated Nature paper, which says that pre-trained models used in the paper can be found at this repository. Can you point me to where I can find these as well as an example of the input data so that I can give the model a try.

Hi,

Have you ordered the ~3000 1-Mbp bins by genomic coordinates and chromosome number (chr1, chr2, etc)?

I want to use the mutational density values that you provided but I am unsure whether the bins are ordered (bin1: chr1:1000000-1999999, bin2: chr1:200000-2999999) or randomly shuffled.

Thank you!

I read your paper 'A deep learning system accurately classifies primary and metastatic cancers using passenger mutation patterns'. I would like to reproduce the result from this paper by training the random forest and neural network model you provided here.
I would be glad if you could provide me the value of each feature (that you mentioned in Table 2: WGS feature types used in classifier) for PCAWG training and test datasets in a format that can directly be used to train and test the models. The feature vectors for the independent validation dataset would also be very helpful.

I read your amazing nature paper and would like to try out the DNN method. Could you please provide a more detailed guideline about how to use your code, such as giving some examples to reproduce the results in the paper? Specifically, I have following questions. I wonder what the feature_type could be? Regarding the Keras model in the model folder, I wonder which folder you used? Could you please also provide the codes that how you slipt the training, validation, and testing samples? Thanks in advance.

Hello, I would also like to run this model. Could you please refer me to the place you mentioned the input file format for predict_cancer.py?