multiple updates to CLI (#331)

* multiple updates to CLI

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

geneplexus/cli.py

@@ -1,6 +1,7 @@
 """Command line interface for the GenePlexus pipeline."""
 import argparse
 import atexit
+import json
 import os
 import os.path as osp
 import pathlib
@@ -37,7 +38,7 @@ def parse_args() -> argparse.Namespace:
         "--input_file",
         metavar="",
         required=True,
-        help="Input gene list (.txt) file (one gene per line).",
+        help="Input gene list (.txt) file.",
     )
 
     parser.add_argument(
@@ -50,6 +51,15 @@ def parse_args() -> argparse.Namespace:
         "tabs. Other generic separator are also supported, e.g. ', '.",
     )
 
+    parser.add_argument(
+        "-dd",
+        "--data_dir",
+        default=None,
+        metavar="",
+        help="Directory in which the data are stored, if set to None, then use "
+        "the default data directory ~/.data/geneplexus",
+    )
+
     parser.add_argument(
         "-n",
         "--network",
@@ -72,24 +82,31 @@ def parse_args() -> argparse.Namespace:
         "--sp_trn",
         default="Human",
         metavar="",
-        help="Species of training data {format_choices(config.ALL_SPECIES}",
+        help=f"Species of training data {format_choices(config.ALL_SPECIES)}",
     )
 
     parser.add_argument(
         "-s2",
         "--sp_tst",
-        default="Human",
+        default="Mouse",
+        metavar="",
+        help=f"Species of test data {format_choices(config.ALL_SPECIES)}",
+    )
+
+    parser.add_argument(
+        "-g1",
+        "--gsc_trn",
+        default="GO",
         metavar="",
-        help="Species of test data {format_choices(config.ALL_SPECIES}",
+        help=f"Geneset collection used to generate negatives. {format_choices(config.ALL_GSCS)}",
     )
 
     parser.add_argument(
-        "-g",
-        "--gsc",
+        "-g2",
+        "--gsc_tst",
         default="GO",
         metavar="",
-        help="Geneset collection used to generate negatives and the model"
-        f"similarities. {format_choices(config.ALL_GSCS)}",
+        help=f"Geneset collection used for model similarities. {format_choices(config.ALL_GSCS)}",
     )
 
     parser.add_argument(
@@ -101,15 +118,6 @@ def parse_args() -> argparse.Namespace:
         help="Number of nodes in the small edgelist.",
     )
 
-    parser.add_argument(
-        "-dd",
-        "--data_dir",
-        default=None,
-        metavar="",
-        help="Directory in which the data are stored, if set to None, then use "
-        "the default data directory ~/.data/geneplexus",
-    )
-
     parser.add_argument(
         "-od",
         "--output_dir",
@@ -126,6 +134,13 @@ def parse_args() -> argparse.Namespace:
         help=f"Logging level. {format_choices(config.LOG_LEVELS)}",
     )
 
+    parser.add_argument(
+        "-ad",
+        "--auto_download_off",
+        action="store_true",
+        help="Turns off autodownloader which is on by default.",
+    )
+
     parser.add_argument(
         "-q",
         "--quiet",
@@ -155,15 +170,19 @@ def parse_args() -> argparse.Namespace:
     parser.add_argument(
         "--skip-mdl-sim",
         action="store_true",
-        help="Skip model similarity computation. This computation is not yet "
-        "available when using custom networks due to the lack of pretrained "
-        "models for comparison.",
+        help="Skip model similarity computation",
+    )
+
+    parser.add_argument(
+        "--skip-sm-edgelist",
+        action="store_true",
+        help="Skip making small edgelist.",
     )
 
     return parser.parse_args()
 
 
-def run_pipeline(gp: GenePlexus, num_nodes: int, skip_mdl_sim: bool):
+def run_pipeline(gp: GenePlexus, num_nodes: int, skip_mdl_sim: bool, skip_sm_edgelist: bool):
     """Run the full GenePlexus pipeline.
 
     Args:
@@ -175,12 +194,15 @@ def run_pipeline(gp: GenePlexus, num_nodes: int, skip_mdl_sim: bool):
 
     """
     gp.fit_and_predict()
-    gp.make_small_edgelist(num_nodes=num_nodes)
-    gp.alter_validation_df()
     if not skip_mdl_sim:
         gp.make_sim_dfs()
     else:
         logger.info("Skipping model similarity computation.")
+    if not skip_sm_edgelist:
+        gp.make_small_edgelist(num_nodes=num_nodes)
+    else:
+        logger.info("Skipping making small edgelist.")
+    gp.alter_validation_df()
 
 
 def df_to_tsv(df: pd.DataFrame, root: str, name: str):
@@ -195,29 +217,37 @@ def df_to_tsv(df: pd.DataFrame, root: str, name: str):
     df.to_csv(osp.join(root, name), sep="\t", index=False)
 
 
-def save_results(gp, outdir, zip_output, overwrite, skip_mdl_sim):
+def save_results(gp, outdir, zip_output, overwrite, skip_mdl_sim, skip_sm_edgelist):
     """Save all results generated by the GenePlexus pipeline.
 
     Args:
         outdir: Output directory.
         zip_output: Whether or not to zip the output directory into a zip file.
         overwrite: Whether or not to overwrite existing results.
         skip_mdl_sim: Whether or not to skip the computation of model
-            similarities with GO and Mondo. This option is not yet available
-            for custom networks.
+            similarities with GO, Monarch and/or Mondo.
+        skip_sm_edgelist: Whether or not to skip making the small edgelist.
 
     """
     zip_outpath = _suffix_fn(f"{outdir}.zip", overwrite=overwrite)
     outdir = _suffix_dir(outdir, overwrite=overwrite, mktmp=zip_output)
-
-    np.savetxt(osp.join(outdir, "cross_validation.txt"), gp.avgps, fmt="%.18f")
     df_to_tsv(gp.df_convert_out, outdir, "df_convert_out.tsv")
+    np.savetxt(osp.join(outdir, "pos_genes_in_net.txt"), gp.pos_genes_in_net, fmt="%s")
+    np.savetxt(osp.join(outdir, "negative_genes.txt"), gp.negative_genes, fmt="%s")
+    np.savetxt(osp.join(outdir, "net_genes.txt"), gp.net_genes, fmt="%s")
+    with open(osp.join(outdir, "neutral_gene_info.json"), "w") as f:
+        json.dump(gp.neutral_gene_info, f)
+    np.savetxt(osp.join(outdir, "avgps.txt"), gp.avgps, fmt="%.18f")
+    np.savetxt(osp.join(outdir, "mdl_weights.txt"), gp.mdl_weights, fmt="%.18f")
     df_to_tsv(gp.df_probs, outdir, "df_probs.tsv")
-    df_to_tsv(gp.df_edge, outdir, "df_edge.tsv")
-    df_to_tsv(gp.df_edge_sym, outdir, "df_edge_sym.tsv")
-    df_to_tsv(gp.df_convert_out_subset, outdir, "df_convert_out_subset.tsv")
     if not skip_mdl_sim:
         df_to_tsv(gp.df_sim, outdir, "df_sim.tsv")
+    if not skip_mdl_sim:
+        df_to_tsv(gp.df_edge, outdir, "df_edge.tsv")
+        df_to_tsv(gp.df_edge_sym, outdir, "df_edge_sym.tsv")
+        np.savetxt(osp.join(outdir, "isolated_genes.txt"), gp.isolated_genes, fmt="%s")
+        np.savetxt(osp.join(outdir, "isolated_genes_sym.txt"), gp.isolated_genes_sym, fmt="%s")
+    df_to_tsv(gp.df_convert_out_subset, outdir, "df_convert_out_subset.tsv")
 
     # Dump config, close file handler and move run log to result directory
     gp.dump_config(outdir)
@@ -302,18 +332,23 @@ def main():
     """Command line interface."""
     args = parse_args()
     log_level = "CRITICAL" if args.quiet else args.log_level
+    if args.auto_download_off:
+        auto_download = False
+    else:
+        auto_download = True
 
     clear_data(args)
 
     # Create geneplexus object and auto download data files
     gp = GenePlexus(
-        args.data_dir,
-        args.network,
-        args.feature,
-        args.sp_trn,
-        args.sp_tst,
-        args.gsc,
-        auto_download=True,
+        file_loc=args.data_dir,
+        net_type=args.network,
+        features=args.feature,
+        sp_trn=args.sp_trn,
+        sp_tst=args.sp_tst,
+        gsc_trn=args.gsc_trn,
+        gsc_tst=args.gsc_tst,
+        auto_download=auto_download,
         log_level=log_level,
     )
 
@@ -323,8 +358,10 @@ def main():
     # Save config
 
     # Run pipeline and save results
-    run_pipeline(gp, args.small_edgelist_num_nodes, args.skip_mdl_sim)
-    save_results(gp, normexpand(args.output_dir), args.zip_output, args.overwrite, args.skip_mdl_sim)
+    run_pipeline(gp, args.small_edgelist_num_nodes, args.skip_mdl_sim, args.skip_sm_edgelist)
+    save_results(
+        gp, normexpand(args.output_dir), args.zip_output, args.overwrite, args.skip_mdl_sim, args.skip_sm_edgelist
+    )
 
 
 if __name__ == "__main__":