Merge pull request #46 from aws-solutions-library-samples/v1.13.2

V1.13.2

Author: anindith123

CHANGELOG.md

@@ -7,11 +7,17 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ---
 
-## [1.13.2] - 2025-06-10
+## [1.13.2] - 2025-09-10
 
 ### 1.13.2 Fixed
 
 - Updated build artifact location
+- Fix issue limiting deployment to us-east-1 only
+- Update SageMaker notebook platform ID
+- Fix issue downloading ESMFold parameters
+- Fix multiple issues with ESMFold inference script
+- Fix issue with version of numpy in RFDiffusion ontainer
+
 
 ---
 

README.md

@@ -1,5 +1,11 @@
 # AWS Batch Architecture for Protein Folding and Design
 
+> [!IMPORTANT]
+> The examples provided in this repository are for experimental and educational purposes only. They demonstrate concepts and techniques but are not intended for direct use in production environments.
+
+> [!IMPORTANT]
+> The CloudFormation templates included in this repository only support deployment to the us-east-1, and us-east-2, eu-west-1, eu-west-2, eu-central-1, and ap-southwest-1 regions. For additional options, please refer to the [Drug Discovery Workflows for AWS HealthOmics](https://github.com/aws-samples/drug-discovery-workflows) project.
+
 ## Contents
 
 1. [Overview](#1-overview)  

infrastructure/cloudformation/batch-protein-folding-cfn-notebook.yaml

@@ -123,7 +123,7 @@ Resources:
       InstanceType: ml.t3.medium
       DefaultCodeRepository: !Ref CodeRepositoryURI
       KmsKeyId: !GetAtt EncryptionKey.Arn
-      PlatformIdentifier: "notebook-al2-v2"
+      PlatformIdentifier: "notebook-al2-v3"
       RoleArn: !GetAtt SageMakerNotebookExecutionRole.Arn
       SubnetId: !Ref Subnet
       SecurityGroupIds:

infrastructure/docker/alphafold2/Dockerfile

@@ -15,16 +15,16 @@ ARG ALPHAFOLD2_VERSION
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
 RUN apt-get update \ 
-    && DEBIAN_FRONTEND=noninteractive apt-get install --no-install-recommends -y \
-    build-essential \
-    cmake \
-    cuda-command-line-tools-$(cut -f1,2 -d- <<< ${CUDA//./-}) \
-    git \
-    hmmer \
-    kalign \
-    tzdata \
-    wget \
-    awscli \
+  && DEBIAN_FRONTEND=noninteractive apt-get install --no-install-recommends -y \
+  build-essential \
+  cmake \
+  cuda-command-line-tools-$(cut -f1,2 -d- <<< ${CUDA//./-}) \
+  git \
+  hmmer \
+  kalign \
+  tzdata \
+  wget \
+  awscli \
   && rm -rf /var/lib/apt/lists/* \
   && apt-get autoremove -y \
   && apt-get clean
@@ -40,43 +40,39 @@ RUN git clone --branch v3.3.0 https://github.com/soedinglab/hh-suite.git /tmp/hh
   && rm -rf /tmp/hh-suite
 
 # Install Miniconda package manager.
-RUN wget -q -P /tmp \
-    https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh \ 
-  && bash /tmp/Miniconda3-latest-Linux-x86_64.sh -b -p /opt/conda \
-  && rm /tmp/Miniconda3-latest-Linux-x86_64.sh
+RUN wget -q -P /tmp \    
+  "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-x86_64.sh" \
+  && bash /tmp/Miniforge3-Linux-x86_64.sh -b -p /opt/conda \
+  && rm /tmp/Miniforge3-Linux-x86_64.sh  
 
 # Install conda packages.
 ENV PATH="/opt/conda/bin:$PATH"
-# RUN conda install -qy conda==4.13.0 
-#   && conda install -y -c conda-forge
-RUN conda install -y -c conda-forge \
-    openmm=7.5.1 \    
-    cudatoolkit=${CUDA_VERSION} \
-    pdbfixer=1.7 \
-    pip \
-    python=3.9.16 \
+ENV LD_LIBRARY_PATH="/opt/conda/lib:$LD_LIBRARY_PATH"
+
+RUN conda install --quiet --yes --channel conda-forge pip python=3.11 openmm=8.0.0 pdbfixer --override-channels \
+  && conda install --quiet --yes --channel nvidia cuda=${CUDA_VERSION} --override-channels \
   && conda clean --all --force-pkgs-dirs --yes
 
-RUN git clone --branch ${ALPHAFOLD2_VERSION} --depth 1 https://github.com/deepmind/alphafold.git /app/alphafold
+# RUN git clone --branch ${ALPHAFOLD2_VERSION} --depth 1 https://github.com/deepmind/alphafold.git /app/alphafold
+RUN git clone --depth 1 https://github.com/deepmind/alphafold.git /app/alphafold
 
 RUN wget -q -P /app/alphafold/alphafold/common/ \
-    https://git.scicore.unibas.ch/schwede/openstructure/-/raw/7102c63615b64735c4941278d92b554ec94415f8/modules/mol/alg/src/stereo_chemical_props.txt
+  https://git.scicore.unibas.ch/schwede/openstructure/-/raw/7102c63615b64735c4941278d92b554ec94415f8/modules/mol/alg/src/stereo_chemical_props.txt
 
 # Install pip packages.
 RUN pip3 install --upgrade pip --no-cache-dir \
-    && pip3 install -r /app/alphafold/requirements.txt --no-cache-dir \
-    && pip3 install --upgrade --no-cache-dir \
-      jax==0.3.25 \
-      jaxlib==0.3.25+cuda11.cudnn805 \
-      -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
-
-# Apply OpenMM patch.
-WORKDIR /opt/conda/lib/python3.9/site-packages
-RUN patch -p0 < /app/alphafold/docker/openmm.patch
+  && pip3 install -r /app/alphafold/requirements.txt --no-cache-dir \
+  && pip3 install --upgrade --no-cache-dir \
+  jax==0.4.26 \
+  jaxlib==0.4.26+cuda12.cudnn89 \
+  -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
 
 # Add SETUID bit to the ldconfig binary so that non-root users can run it.
 RUN chmod u+s /sbin/ldconfig.real
 
+# Currently needed to avoid undefined_symbol error.
+RUN ln -sf /usr/lib/x86_64-linux-gnu/libffi.so.7 /opt/conda/lib/libffi.so.7
+
 # We need to run `ldconfig` first to ensure GPUs are visible, due to some quirk
 # with Debian. See https://github.com/NVIDIA/nvidia-docker/issues/1399 for
 # details.
@@ -85,8 +81,8 @@ RUN chmod u+s /sbin/ldconfig.real
 WORKDIR /app/alphafold
 COPY run.sh /app/alphafold/run.sh
 RUN echo $'#!/bin/bash\n\
-ldconfig\n\
-python /app/alphafold/run_alphafold.py "$@"' > /app/run_alphafold.sh \
+  ldconfig\n\
+  python /app/alphafold/run_alphafold.py "$@"' > /app/run_alphafold.sh \
   && chmod +x /app/run_alphafold.sh /app/alphafold/run.sh
 
 ENTRYPOINT ["bash", "/app/alphafold/run.sh"]

infrastructure/docker/download/Dockerfile

@@ -5,10 +5,12 @@ FROM public.ecr.aws/amazonlinux/amazonlinux:2
 COPY scripts /scripts
 RUN amazon-linux-extras install epel -y \
   && yum update -y \
-  && yum install aria2 tar rsync unzip -y \
+  && yum install aria2 tar rsync unzip git git-lfs -y \
   && curl  "https://awscli.amazonaws.com/awscli-exe-linux-$(uname -m).zip" -o "awscliv2.zip" \
   && unzip awscliv2.zip \
   && ./aws/install \
   && chmod -R 755 /scripts
 
+WORKDIR /home/scripts
+
 ENTRYPOINT ["bash", "-c"]

infrastructure/docker/download/scripts/download_esmfold_params.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 #
 # Original Copyright 2022 AlQuraishi Laboratory
-# Modifications Copyright 2023 Amazon.com, Inc. or its affiliates. All Rights Reserved.
+# Modifications Copyright 2025 Amazon.com, Inc. or its affiliates. All Rights Reserved.
 # SPDX-License-Identifier: Apache-2.0
 #
 # Downloads and unzips the ESMFold parameters.
@@ -19,9 +19,10 @@ ROOT_DIR="${DOWNLOAD_DIR}/esmfold_params/hub"
 SOURCE_URL="facebook/esmfold_v1"
 
 git lfs install
-echo "Downloading ${SOURCE_URI} from Hugging Face Hub"
+echo "Downloading ${SOURCE_URL} from Hugging Face Hub"
 mkdir tmp
-git clone https://huggingface.co/$SOURCE_URI tmp --depth=1
+git clone https://huggingface.co/$SOURCE_URL tmp --depth=1
 rm -rf tmp/.git
+mkdir -p $ROOT_DIR
 mv -n tmp/* $ROOT_DIR
 rm -rf tmp

infrastructure/docker/esmfold/Dockerfile

@@ -3,10 +3,17 @@ FROM nvcr.io/nvidia/cuda:12.6.3-runtime-amzn2023 as base
 WORKDIR /home
 
 RUN dnf update -y \
-    && dnf install -y 'python3.11' \
+    && dnf install -y 'python3.11' unzip which \
     && dnf clean all \
     && rm -rf /var/cache/dnf/*
 
+# Install AWS CLI
+RUN curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip" \ 
+    && unzip awscliv2.zip \
+    && ./aws/install \
+    && rm -rf aws \
+    && rm awscliv2.zip
+
 ENV VIRTUAL_ENV='/opt/venv'
 ENV PATH="${VIRTUAL_ENV}/bin:$PATH"
 

infrastructure/docker/esmfold/scripts/esmfold_inference.py

@@ -66,7 +66,7 @@ def predict_structures(
 
         output = {key: value.cpu() for key, value in outputs.items()}
         pdb_string = model.output_to_pdb(output)[0]
-        output_dir = os.path.join(args.output_dir, str(n))
+        output_dir = os.path.join(output_dir, str(n))
         if not os.path.exists(output_dir):
             os.makedirs(output_dir)
 

infrastructure/docker/rfdiffusion/Dockerfile

@@ -26,6 +26,7 @@ RUN apt-get -q update \
   hydra-core==1.3.2 \
   pyrsistent==0.19.3 \
   icecream==2.1.3 \
+  numpy==1.26.4 \
   /app/RFdiffusion/env/SE3Transformer \
   && wget -O "awscliv2.zip" "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" \
   && unzip awscliv2.zip \

src/batchfold/esmfold_job.py

@@ -29,11 +29,11 @@ def __attrs_post_init__(self) -> None:
         command_list = [f"-i {self.fasta_s3_uri}:input/{self.target_id}.fasta"]
         command_list.extend([f"-o output/:{self.output_s3_uri}"])
         command_list.extend([
-            "python scripts/esmfold_inference.py",
-            f"--fasta=input/{self.target_id}.fasta",
+            "python esmfold_inference.py",
             f"--pdb=output",
             f"--num-recycles={self.num_recycles}",
             f"--model-dir={self.model_dir}",
+            f"input/{self.target_id}.fasta",
         ])
 
         if self.chunk_size: