Idgl branch

.gitignore

@@ -165,3 +165,52 @@ examples/multi_modality/joint_embedding/models/*.pth
 trash
 out.log
 temp_data
+examples/evolo/environment.sh
+examples/evolo/**/*.json
+examples/evolo/**/*.png
+examples/evolo/**/test/
+examples/evolo/**/train/
+examples/search/**/*.json
+Image_crop
+test
+examples/evolo/cta_scgat/map/human/map.xlsx
+examples/evolo/**/*/old/
+examples/evolo/**/*/spatial/
+examples/graph_automl/**/*/train/
+examples/graph_automl/**/*/test/
+examples/graph_automl/**/*/spatial/
+examples/search/**/*/logs/
+examples/similarity/GraphEvolve/openevolve
+/mnt/nfs/zyxing/msu/dance_temp/dance/examples/similarity/GraphEvolve/test_db
+/mnt/nfs/zyxing/msu/dance_temp/dance/examples/similarity/GraphEvolve/test_db2
+/mnt/nfs/zyxing/msu/dance_temp/dance/examples/similarity/GraphEvolve/test_db3
+/mnt/nfs/zyxing/msu/dance_temp/dance/examples/similarity/GraphEvolve/test_db4
+examples/evolo/**/*/map
+
+dance/metadata/STRINGDB.graph.csv
+chroma_server_data
+temp.py
+temp.csv
+examples/graph_automl/**/*/map
+old
+examples/similarity/GraphEvolve/train
+examples/similarity/GraphEvolve/data/spatial
+examples/similarity/GraphEvolve/map
+examples/single_modality/cell_type_annotation/output/models/best_model.pth
+examples/single_modality/cell_type_annotation/output/models/final_model.pth
+map/celltype2subtype.xlsx
+map/human/map.xlsx
+map/mouse/map.xlsx
+train/human/human_Brain328_celltype.csv
+train/human/human_Brain328_data.csv
+examples/graph_automl/cta_scrgcl/temp_output_2/models/best_model.pth
+examples/graph_automl/cta_scrgcl/temp_output_2/models/final_model.pth
+examples/graph_automl/cta_scrgcl/temp_output_0/models/best_model.pth
+examples/graph_automl/cta_scrgcl/temp_output_0/models/final_model.pth
+examples/graph_automl/cta_scrgcl/temp_output_1/models/best_model.pth
+examples/graph_automl/cta_scrgcl/temp_output_1/models/final_model.pth
+examples/evolo/cta_scrgcl/oo.csv
+examples/evolo/enviroment.sh
+examples/search/generate_pseudocode/.deepeval/.deepeval_telemetry.txt
+examples/idgl/cta_scgat/train
+examples/idgl/cta_scgat/map

README.md

@@ -5,6 +5,22 @@
   />
 </p>
 
+# CEAgent-GSL: Code-level Evolutionary Agent for Interpretable Graph Structure Learning on Omics
+
+> **Note to Reviewers and Users:**
+> This repository contains the official implementation of **ECAgent-GSL**. While **partially utilizing** the **DANCE 2.0** environment for infrastructure support, our contributions are concentrated in specific modules tailored for graph topology learning. The core contributions of this work are independent and strictly isolated.
+>
+> **Key Implementation Directories:**
+>
+> - **`examples/evolo/`**: The core evolutionary loop and LLM agent interaction logic.
+> - **`examples/search/` & `examples/similarity/`**: The Intelligent Recommendation and retrieval modules.
+> - **`dance/modules/`**: A **curated benchmark suite**. From the original ~30 methods, we selectively retained **5 legacy algorithms** relevant to our task and **implemented 6 new graph-based algorithms from scratch**. This results in a specialized collection of 11 baselines strictly controlled for our experiments.
+> - **`dance/transforms/graph/`**: An existing module **significantly expanded** in this work. We injected diverse graph construction logic (extracted from the 11 baselines) into this module, transforming it into the primary **search space** for our evolutionary optimization.
+>
+> The remaining parts of this repository serve as the environmental substrate for data loading and reproducibility.
+
+______________________________________________________________________
+
 ______________________________________________________________________
 
 [![PyPI version](https://badge.fury.io/py/pydance.svg)](https://badge.fury.io/py/pydance)