Instructions to use Shenzhi-Chen/DeepSTARR-Mouse with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use Shenzhi-Chen/DeepSTARR-Mouse with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://Shenzhi-Chen/DeepSTARR-Mouse") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -34,10 +34,15 @@ The code and architecture definitions are available on **GitHub:**
|
|
| 34 |
|
| 35 |
**Model Weights**
|
| 36 |
**Sequence‑to‑accessibility** and **sequence‑to‑activity** model weights are stored separately in two folders.
|
| 37 |
-
Each folder contains
|
| 38 |
Within each tissue folder there are **six models**, corresponding to **three cross‑validation folds** and **two replicates**.
|
| 39 |
Model weights are stored in **Model.json** (architecture) and **Model.h5** (trained parameters).
|
| 40 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
---
|
| 42 |
|
| 43 |
**Training Objectives and Evaluation Metrics**
|
|
|
|
| 34 |
|
| 35 |
**Model Weights**
|
| 36 |
**Sequence‑to‑accessibility** and **sequence‑to‑activity** model weights are stored separately in two folders.
|
| 37 |
+
Each folder contains six tissue‑specific subfolders: **heart**, **limb**, **midbrain (CNS)**, **forebrain**, **hindbrain** and **neural tube**
|
| 38 |
Within each tissue folder there are **six models**, corresponding to **three cross‑validation folds** and **two replicates**.
|
| 39 |
Model weights are stored in **Model.json** (architecture) and **Model.h5** (trained parameters).
|
| 40 |
|
| 41 |
+
**K27Ac_K4me1_models**, **promoter_distal_no_CTCF_models** and **tissueSpecATAC** extended VISTA replacement transfer learning model weights are stored separately in three folders.
|
| 42 |
+
Each folder contains three tissue‑specific subfolders: **heart**, **limb**, **midbrain**
|
| 43 |
+
Within each tissue folder there are **six models**, corresponding to **three cross‑validation folds** and **two replicates**.
|
| 44 |
+
Model weights are stored in **Model.json** (architecture) and **Model.h5** (trained parameters).
|
| 45 |
+
|
| 46 |
---
|
| 47 |
|
| 48 |
**Training Objectives and Evaluation Metrics**
|