Upload 3 files

DNABERT2-FINAL.py

@@ -0,0 +1,67 @@
+
+import streamlit as st
+from transformers import AutoTokenizer, AutoModelForSequenceClassification, BertConfig
+import torch
+
+
+model_name = "rashiqua/dnabert2_epigenetic"
+
+
+config = BertConfig.from_pretrained(model_name)
+
+
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+
+
+model = AutoModelForSequenceClassification.from_pretrained(model_name, trust_remote_code=True, config=config)
+
+def main():
+    st.title("Epigenetic Marks Prediction")
+    st.write("An application of DNA BERT2")
+
+
+    st.sidebar.header("About")
+    st.sidebar.write("This app uses DNA BERT2 to predict the presence of epigenetic marks in a given DNA sequence.")
+
+
+    user_input = st.text_area("Enter a DNA sequence:", height=150)
+
+    if st.button("Classify Sequence"):
+        if user_input:
+
+            predicted_class, confidence = pred(user_input)
+
+
+            st.subheader("Prediction Result")
+            if predicted_class == 1:
+                st.success("Epigenetic Mark detected!")
+            else:
+                st.info("No epigenetic mark found.")
+
+            st.subheader("Class Distribution")
+            st.write("1 - Epigenetic mark found")
+            st.progress(confidence)
+            st.text(f"{confidence * 100:.2f}%")
+            
+            st.write("0 - Epigenetic mark not found")
+            st.progress(1 - confidence)
+            st.text(f"{(1 - confidence) * 100:.2f}%")
+
+        else:
+            st.warning("Please enter a DNA sequence for classification.")
+
+
+def pred(sequence):
+    encoded_input = tokenizer(sequence, return_tensors='pt')
+    
+    with torch.no_grad():
+        outputs = model(input_ids=encoded_input['input_ids'], attention_mask=encoded_input['attention_mask'])
+        logits = outputs[0]
+        predicted_class = logits.argmax(-1).item()
+        confidence = logits.softmax(dim=-1)[0, 1].item()
+
+    return predicted_class, confidence
+
+
+if __name__ == "__main__":
+    main()

Dockerfile

@@ -0,0 +1,24 @@
+FROM python:3.12
+
+RUN useradd -m -u 1000 user
+
+USER user
+
+ENV HOME=/home/user \
+    PATH=/home/user/.local/bin:$PATH
+
+WORKDIR $HOME/app
+
+RUN pip install --no-cache-dir --upgrade pip
+
+COPY --chown=user . $HOME/app
+
+RUN pip install --user -r requirements.txt
+
+RUN pip uninstall -y triton
+
+RUN mkdir -p $HOME/.cache/huggingface && chmod 777 $HOME/.cache/huggingface
+
+EXPOSE 7860
+
+CMD ["streamlit", "run", "DNABERT2-FINAL.py", "--server.port=7860", "--server.address=0.0.0.0"]

requirements.txt

@@ -0,0 +1,9 @@
+einops
+transformers
+peft
+omegaconf
+torch
+evaluate
+accelerate
+streamlit
+