Example and tutorial data for PHILM

The data are used in Github Readme and PHILM Step-by-Step Tutorial. The provided archives include: * `example.zip`: Example dataset used in the README. * `tutorial_data.zip`: Healthy human stool sample dataset used in the step-by-step tutorial. * `PHILM_input_data.7z`: Contains the input files required by PHILM. These files should be placed in the `data/` directory. * `PHILM_model_results.7z`: Contains the trained model and prediction metrics, including `PHILM_best_model.pth`, `PHILM_best_params.yaml`, `PHILM_predict_test.ft.metrics`, `PHILM_predict_val.ft.metrics`, and `PHILM_predict_train.ft.metrics`. * `PHILM_interactions.7z`: Contains the inferred interaction results, including `PHILM_interactions.tsv`, `PHILM_interactions.pvalues.tsv`, and `PHILM_interactions.pvalues.filtered.tsv`. * `perm_<start number-end number>_scores.tsv.7z`: * `perm_<start-end>_scores.tsv.7z`: These archives contain the permutation-derived PHILM scores used for empirical p-value calculation. Because storing all 1,000 permutation results in a single file would be very large, the results were divided into 50 compressed files. After decompressing all `.7z` files, organize the permutation results into the required PHILM directory structure by running: `python organize_perm_files.py --input "perm_*-*_scores.tsv" --outdir permutation_null`. This command will generate files with the following structure: `permutation_null/perm_*/PHILM_perm_*.raw_gradient.tsv`. Before running Step 5.3, move or place the generated `permutation_null/` directory under the `results/` directory.