initial commit

.gitattributes

@@ -56,3 +56,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+human_methylation_bench_ver1_adata_test.h5ad filter=lfs diff=lfs merge=lfs -text
+human_methylation_bench_ver1_main_test.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,46 @@
----
-license: cc-by-4.0
----
+---
+license: cc-by-4.0
+language:
+  - en
+tags:
+  - biology
+  - Aging
+
+configs:
+  - config_name: default
+    data_files:
+      - split: test
+        path: "human_methylation_bench_ver1_main_test.csv"
+---
+
+# Human DNA Methylation Dataset ver1
+
+This dataset is a benchmark dataset for predicting the aging clock, curated from publicly available DNA methylation data. The original benchmark dataset was published by [Dmitrii Kriukov et al. (2024)](https://www.biorxiv.org/content/10.1101/2024.06.06.597715v1.full) by integrating data from 65 individual studies.
+
+To improve usability, we ensured unique sample IDs (excluding duplicate data, GSE118468 and GSE118469) and randomly split the data into training and testing subsets (train : test = 7 : 3) to facilitate model evaluation.
+
+## Data
+
+Methylation profiles (stored as `main.csv`), containing CpG beta values across samples. Total : 10,386 samples and 13,655 CpG sites.
+
+- **`main`**: Methylation data table, containing beta values per CpG site for each sample.
+- **`meta`**: Metadata table, detailing sample information such as `SampleID`, `DatasetID`, `PlatformID`, `Tissue`, `CellType`, `Gender`, `Age`, `Condition`, `Class`.
+- **`adata`**: Advanced format for users requiring deeper analysis.
+
+## Files
+
+- `human_methylation_bench_ver1_main_train.csv`, `human_methylation_bench_ver1_main_test.csv`
+- `human_methylation_bench_ver1_meta_train.csv`, `human_methylation_bench_ver1_meta_test.csv`
+- `human_methylation_bench_ver1_adata_train.h5ad`, `human_methylation_bench_ver1_adata_test.h5ad`
+
+## Metadata
+
+- **`SampleID`**: Unique identifier for each sample.
+- **`DatasetID`**: GEO ID of the dataset containing the respective sample.
+- **`PlatformID`**: GEO ID of the platform used for sample profiling, which can be mapped to common platform names (e.g., GPL8490 = 27K, GPL13534 = 450K).
+- **`Tissue`**: Source tissue of the sample, such as peripheral blood, saliva, or buccal swab.
+- **`CellType`**: Type of cells in the sample, which could be a specific cell population (e.g., immune cell subtypes) or broader categories like whole blood, buffy coat, or PBMC.
+- **`Gender`**: Donor gender (M = Male, F = Female, O = Other, U = Unknown).
+- **`Age`**: Chronological age of the donor in years, either rounded or calculated from smaller units.
+- **`Condition`**: Health or disease status of the donor, used to distinguish between healthy controls (HC) and aging-accelerating conditions (AACs). AACs include Atherosclerosis (AS), Ischemic Heart Disease (IHD), Cerebrovascular Accident (CVA), Inflammatory Bowel Disease (IBD), Human Immunodeficiency Virus infection (HIV), Extreme Obesity (XOB, defined by BMI ≥ 40 kg/m²), Type 1 Diabetes Mellitus (T1D), Type 2 Diabetes Mellitus (T2D), Rheumatoid Arthritis (RA), Osteoporosis (OP), Alzheimer's Disease (AD), Parkinson's Disease (PD), Multiple Sclerosis (MS), Creutzfeldt-Jakob Disease (CJD), Chronic Obstructive Pulmonary Disease (COPD), Tuberculosis (TB), Werner Syndrome (WS), Hutchinson-Gilford Progeria Syndrome (HGPS), and Congenital Generalized Lipodystrophy (CGL).
+- **`Class`**: Class of the sample's condition, where healthy controls are in a separate class labeled as HC.

human_methylation_bench_ver1_adata_test.h5ad

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4ae00418d31d8dd1d34fc688311a7aacfd42defe679c496f14ea67d7f38a4ed
+size 171104220

human_methylation_bench_ver1_main_test.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b7d4093e411a980a787ec2449b66036d9fd6bd8c94cb419f4877e79951b2f36
+size 409000040