feat: start working on docker

Dockerfile

@@ -0,0 +1,54 @@
+# Base image containing the installed gt4sd environment
+FROM gt4sd-base:latest
+
+
+# Clone this repository
+RUN git clone https://github.com/GT4SD/molecular-design.git
+
+WORKDIR /molecular-design
+
+# We need to use the pypi toxsmi package, not the default one
+RUN pip uninstall --yes toxsmi && \
+    pip install toxsmi
+
+
+# Load datasets
+RUN mkdir data && \
+    python gt4sd-core/scripts/load_data.py \
+      --uniprot P05067 \
+      --affinity_type IC50 \
+      --affinity_cutoff 10000 \
+      --output_dir data/ \
+      --train_size 0.8 \
+      --binary_labels
+
+# Train toxsmi model
+RUN python gt4sd-core/scripts/train_toxsmi.py \
+      --train data/train.csv \
+      --test data/valid.csv \
+      --smi data/mols.smi \
+      --language tokenizer \
+      --params gt4sd-core/config/toxsmi_conf.json \
+      --model models \
+      --name toxsmi_model
+
+# Generate molecules with MoLeR
+RUN head -n 5 data/mols.smi > data/good_docks.smi && \
+    python gt4sd-core/scripts/moler_generate_toxsmi.py \
+      --smi_path data/good_docks.smi \
+      --param_path gt4sd-core/config/moler_conf.json \
+      --output_path data/moler_filtered \
+      --predictor_path models/toxsmi_model/weights/best_F1.pt
+
+# Generate more diverse molecules with Regression Transformer
+RUN python gt4sd-core/scripts/prepare_rt_data.py \
+      --smi_path data/moler_filtered/generated.csv \
+      --output_path data/moler_filtered/generated_qed.csv && \
+    head -n 10 data/moler_filtered/generated_qed.csv > data/moler_filtered/generated_qed_selected.csv && \
+    python gt4sd-core/scripts/rt_generate.py \
+      --smi_path data/moler_filtered/generated_qed_selected.csv \
+      --param_path gt4sd-core/config/rt_conf.json \
+      --output_path data/rt
+
+
+CMD ["bash"]