schulter/emogi

An explainable multi-omics graph integration method based on graph convolutional networks to predict cancer genes.

emogi's Issues

Where can I download or find this file, "cases_all_cancers.2018-11-26.json"?

Hi,

I'm trying to reproduce some of the results, but a file and directory specified in Preprocess_CNAs_Firehose.ipynb ("../../data/pancancer/TCGA/mutation/download_new/cases_all_cancers.2018-11-26.json") is missing. Where can I download or find this file?

Thanks

"genemap_search_cancer.txt" and "genemap2.txt"

Could you please elaborate the difference between "genemap_search_cancer.txt" and "genemap2.txt"? I am not sure where I can find the first one

def get_negative_labels(nodes, positives, ppi_network, min_degree=1, verbose=False):
    if verbose:
        print ("{} genes are in network".format(nodes.shape[0]))
    # get rid of the positives (known cancer genes)
    not_positives = nodes[~nodes.Name.isin(positives)]
    if verbose:
        print ("{} genes are in network but not in positives (known cancer genes from NCG)".format(not_positives.shape[0]))

    # get rid of OMIM genes associated with cancer
    omim_cancer_genes = pd.read_csv('../../data/pancancer/OMIM/genemap_search_cancer.txt',
                                    sep='\t', comment='#', header=0, skiprows=3)
    # use fact that nan != nan for filtering out NaN
    sublists = [sublist for sublist in omim_cancer_genes['Gene/Locus'].str.split(',') if sublist == sublist]
    omim_cancer_geneset = [item.strip() for sublist in sublists for item in sublist]
    not_omim_not_pos = not_positives[~not_positives.Name.isin(omim_cancer_geneset)]
    if verbose:
        print ("{} genes are also not in OMIM cancer genes".format(not_omim_not_pos.shape[0]))

    # get rid of all the OMIM disease genes
    omim_genes = pd.read_csv('../../data/pancancer/OMIM/genemap2.txt', sep='\t', comment='#', header=None)
    omim_genes.columns = ['Chromosome', 'Genomic Position Start', 'Genomic Position End', 'Cyto Location',
                        'Computed Cyto Location', 'Mim Number', 'Gene Symbol', 'Gene Name',
                        'Approved Symbol', 'Entrez Gene ID', 'Ensembl Gene ID', 'Comments',
                        'Phenotypes', 'Mouse Gene Symbol/ID']

Data in multiomics_features.tsv is incompatible

Hello, does multiomics_features.tsv contains the node feature matrix? I wonder why the data has no cnv feature and only 15 cancer types instead of 16 per mf/meth/ge.:smiley:

Pytorch Implementation

Hi,
I am currently working on a pytorch implementation of this project which can be found here.

Please feel free to contribute and give feedback.

How to create test_mask and train_mask?

Hello,

It is not really an issue, but I wanted to ask advice on how to best create test_mask and train_mask vectors once you have the training and test datasets ready? Any recommendation?

Thank you!
Sarah

XX not found in gene list. Skipping

I run the LPR script for computing the contributions of my interested genes. But the error was prompted " 'KMT2D' not found in gene list. Skipping". my code is:
python lrp.py -m ./EMOGI/exist_model/EMOGI_STRINGdb -g KMT2D -b False
What is the cause of this?

How to compare methods such as HotNet2 and MutSigCV with those that require label training

How to compare methods such as HotNet2 and MutSigCV with those that require label training？How to calculate AUPR on HotNet2?

lacking a defined variable 'blca_df_reindexed' in preprocess_cnas.ipynb

Hiii ：)
When I ran this code, this was a problem that held me back.
In notebook preprocess_cnas.ipynb, there is one column says:
cna_blca_matrix = cna_sample_matrix.loc[:, cna_sample_matrix.columns.isin(blca_df_reindexed.columns)]
But it turns out that name 'blca_df_reindexed' is not defined.
So what is 'blca_df_reindexed'?
Looking forward to your reply!
Thank!

No such file or directory: '../data/GCN/training'

Hello,

I installed the software and tried to run the toy example, but I get the following error:

No such file or directory: '../data/GCN/training'

There is no "GCN" in the repository, as far as I can see: am I missing something, or where could I get it?

Thank you!

Data Access

Hello,

Can I get an access to the data used for generating positive and negative labels? The vogelstein labels specifically and other positive and negative genes data.

Thanks

Tensorflow

Do we need to have the same version of tensorflow?
I'm getting this error when installing

ERROR: Could not find a version that satisfies the requirement tensorflow==1.15.3 (from versions: 2.5.0, 2.5.1, 2.5.2, 2.5.3, 2.6.0rc0, 2.6.0rc1, 2.6.0rc2, 2.6.0, 2.6.1, 2.6.2, 2.6.3, 2.6.4, 2.7.0rc0, 2.7.0rc1, 2.7.0, 2.7.1, 2.7.2, 2.8.0rc0, 2.8.0rc1, 2.8.0, 2.8.1, 2.9.0rc0, 2.9.0rc1, 2.9.0rc2, 2.9.0)
ERROR: No matching distribution found for tensorflow==1.15.3