alphafold submodule updated and Dockerfile updated.

IntelLabs · Jan 28, 2024 · 469b159 · 469b159
1 parent 5c85ec5
commit 469b159
Show file tree

Hide file tree

Showing 2 changed files with 30 additions and 28 deletions.
diff --git a/applications/alphafold b/applications/alphafold
diff --git a/pipelines/alphafold2-based-protein-folding/Dockerfile b/pipelines/alphafold2-based-protein-folding/Dockerfile
@@ -1,6 +1,7 @@
+ARG FROM_IMAGE=ubuntu:20.04
 # Install Base miniconda image
 ARG BASE_IMAGE=continuumio/miniconda3
-FROM ${BASE_IMAGE}
+FROM ${BASE_IMAGE} as conda_setup
 
 RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
     git build-essential cmake wget tzdata gcc curl gnupg gnupg2 gnupg1 sudo kalign autoconf numactl time vim \
@@ -19,27 +20,26 @@ RUN apt-get update && \
 # Install Anaconda and PIP dependency
 RUN conda update -n base conda
 RUN conda install python==3.9
-RUN conda install -y -c conda-forge openmm pdbfixer
-RUN conda install -y -c bioconda hmmer hhsuite kalign2
-RUN conda install -y jemalloc
-RUN conda install -c intel onednn-cpu-iomp mkl-devel
-
-RUN python -m pip install torch==2.0.1 pybind11
-RUN python -m pip install absl-py biopython chex dm-haiku dm-tree immutabledict jax ml-collections numpy scipy tensorflow pandas psutil tqdm joblib
+RUN conda install -y -c conda-forge openmm=8.0.0=py39h31b9435_1 pdbfixer=1.9=pyh1a96a4e_0
+RUN conda install -y -c bioconda hmmer=3.3.2=h87f3376_2 hhsuite=3.3.0=py39pl5321h67e14b5_5 kalign2=2.04=hec16e2b_3
+RUN conda install -y jemalloc=5.2.1=h6a678d5_6
+RUN conda install -c intel onednn-cpu-iomp=2023.2.0=intel_49516 mkl-devel=2023.2.0=intel_49495
+RUN python -m pip install torch==2.0.1 pybind11==2.11.1
+RUN python -m pip install absl-py==2.0.0 biopython==1.81 chex==0.1.84 dm-haiku==0.0.10 dm-tree==0.1.8 immutabledict==3.0.0  ml-collections==0.1.1 numpy==1.26.1 scipy==1.11.3 tensorflow==2.14.0 pandas==2.1.1 psutil==5.9.6 tqdm==4.65.0 joblib==1.3.2 pragzip==0.6.0
 RUN python -m pip install jax==0.4.21 jaxlib==0.4.21
 RUN python -m pip install intel-extension-for-pytorch==2.0.100
-
-RUN git clone --recursive https://github.com/IntelLabs/Open-Omics-Acceleration-Framework.git
+#COPY . /Open-Omics-Acceleration-Framework/
+RUN git clone https://github.com/IntelLabs/Open-Omics-Acceleration-Framework.git
 RUN cd /Open-Omics-Acceleration-Framework/applications/alphafold
 WORKDIR /Open-Omics-Acceleration-Framework/applications/alphafold
-
+RUN git submodule update --recursive --init -- .
 # Compile HHsuite from source.
 RUN . /opt/intel/oneapi/compiler/latest/env/vars.sh intel64 \
     && . /opt/intel/oneapi/mkl/latest/env/vars.sh intel64 \
     && git clone --recursive https://github.com/IntelLabs/hh-suite.git \
     && cd hh-suite \
     && mkdir build && cd build \
-    && cmake -DCMAKE_INSTALL_PREFIX=`pwd`/release -DCMAKE_CXX_COMPILER="icpx" -DCMAKE_CXX_FLAGS_RELEASE="-O3 -march=icelake-server" .. \
+    && cmake -DCMAKE_INSTALL_PREFIX=`pwd`/release -DCMAKE_CXX_COMPILER="icpx" -DCMAKE_CXX_FLAGS_RELEASE="-O3 -march=native" .. \
     && make -j 4 && make install \
     && ./release/bin/hhblits -h \
     && cd .. && cd .. \
@@ -51,7 +51,7 @@ RUN . /opt/intel/oneapi/compiler/latest/env/vars.sh intel64 \
     && cd hmmer \
     && cp easel_makefile.in easel/Makefile.in \
     && cd easel && autoconf && ./configure --prefix=`pwd` && cd .. \
-    && autoconf && CC=icx CFLAGS="-O3 -march=icelake-server -fPIC" ./configure --prefix=`pwd`/release \
+    && autoconf && CC=icx CFLAGS="-O3 -march=native -fPIC" ./configure --prefix=`pwd`/release \
     && make -j 4 && make install \
     && ./release/bin/jackhmmer -h \
     && cd ..
@@ -61,21 +61,23 @@ RUN cd tpp-pytorch-extension \
     && python -c "from tpp_pytorch_extension.alphafold.Alpha_Attention import GatingAttentionOpti_forward" \
     && cd ..
 
-CMD . /opt/intel/oneapi/compiler/latest/env/vars.sh intel64 \
-    && . /opt/intel/oneapi/mkl/latest/env/vars.sh intel64 \
-    && mkdir weights && mkdir weights/extracted && python extract_params.py --input /data/params/params_model_1.npz --output_dir ./weights/extracted/model_1 \
+
+FROM ${FROM_IMAGE} as builder
+RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+    git build-essential cmake wget tzdata gcc curl gnupg gnupg2 gnupg1 sudo kalign autoconf numactl time vim \
+    && rm -rf /var/lib/apt/lists/* \
+    && apt-get autoremove -y \
+    && apt-get clean
+COPY --from=conda_setup /opt/conda /opt/conda
+COPY  --from=conda_setup /Open-Omics-Acceleration-Framework /Open-Omics-Acceleration-Framework
+WORKDIR /Open-Omics-Acceleration-Framework/applications/alphafold
+
+RUN echo "source /opt/conda/bin/activate " > ~/.bashrc
+CMD source ~/.bashrc
+ENV PATH "/opt/conda/bin:$PATH"
+ENV LD_LIBRARY_PATH "/opt/conda/lib:$LD_LIBRARY_PATH"
+CMD mkdir weights && mkdir weights/extracted && python extract_params.py --input /data/params/params_model_1.npz --output_dir ./weights/extracted/model_1 \
     && python run_multiprocess_pre.py --root_home=/Open-Omics-Acceleration-Framework/applications/alphafold --data_dir=/data --input_dir=/samples --output_dir=/output --model_name=model_1 \
     && LD_PRELOAD=/opt/conda/lib/libiomp5.so:/opt/conda/lib/libjemalloc.so:$LD_PRELOAD \
     MALLOC_CONF="oversize_threshold:1,background_thread:true,metadata_thp:auto,dirty_decay_ms:-1,muzzy_decay_ms:-1" \
     python run_multiprocess_infer.py --root_condaenv=/opt/conda --root_home=/Open-Omics-Acceleration-Framework/applications/alphafold --data_dir=/data --input_dir=/samples --output_dir=/output --model_name=model_1
-    # && time bash online_preproc_baremetal.sh /Open-Omics-Acceleration-Framework/applications/alphafold /data /samples /output \
-    # && time bash online_inference_baremetal.sh /opt/conda /Open-Omics-Acceleration-Framework/applications/alphafold /data /samples /output model_1
-
-
-
-
-
-
-
-
-