JSON output format

The result format

cMonkey-Python stores iteration results in the out directory as JSON files. This has the advantage that there is no assumption about the system that processes the output, JSON is widely supported.

The result format is as described in the following:

{
  "iteration": "<iteration-num>",
  "rows": {"<cluster_num>": ["<row1>", "<row2>", ...], ...},
  "columns": {"<cluster_num>": ["<column1>", "<column2>", ...], ...},
  "residuals": {"<cluster_num>": "<residual>", ...},
  "motifs": {
    "<cluster-num>": {
      "<sequence-type>": [
        {
          "motif-num": <motif-num>,
          "evalue": <e-value>,
          "annotations": [
            {
              "position": <position>,
              "gene": <gene-name>,
              "pvalue": <p-value>,
              "reverse": <true or false>
            }, ...
          ],
          "pssm": [
            [<value1>, ...],
            ...
          ]
        }, ...
      ],
    }, ...
  }
}

The statistics format

Furthermore, cMonkey/Python also saves a normalized gene expression ratios file in tab-separated format and a number of statistics files in JSON format for the defined iterations.

Cluster Viewer is a web based tool that can read and visualize information stored in this format.

The statistics format is as described in the following:

{
  "median-residual": <median-residual>
  "cluster": {
    <cluster-number>: {
      "num_rows": <number of rows>,
      "num_columns": <number of columns>,
      "residual": <residual>
    }, ...
  }
}