Merge branch 'main' into multimer

2023-10-30 16:20:13 -04:00 · 2023-10-30 16:20:13 -04:00 · 5eacd8b689
parent 15850092d3 2338b896c7
commit 5eacd8b689
11 changed files with 65 additions and 165 deletions
--- a/.github/dependabot.yml
+++ b/.github/dependabot.yml
@ -0,0 +1,7 @@
+version: 2
+updates:
+
+  - package-ecosystem: "github-actions"
+    directory: "/"
+    schedule:
+      interval: "daily"
--- a/.github/workflows/docker-image.yml
+++ b/.github/workflows/docker-image.yml
@ -10,6 +10,6 @@ jobs:
  build:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v4
    - name: Build the Docker image
      run: docker build . --file Dockerfile --tag openfold:$(date +%s)
--- a/.github/workflows/undefined_names.yml
+++ b/.github/workflows/undefined_names.yml
@ -4,8 +4,8 @@ jobs:
  undefined_names:
    runs-on: ubuntu-latest
    steps:
-      - uses: actions/checkout@v2
-      - uses: actions/setup-python@v2
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v4
      - run: pip install --upgrade pip
      - run: pip install flake8
      - run: flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics
--- a/13
+++ b/13
@ -1,4 +1,4 @@
-FROM nvidia/cuda:11.3.1-cudnn8-runtime-ubuntu18.04
+FROM nvidia/cuda:11.3.1-cudnn8-devel-ubuntu18.04

 # metainformation
 LABEL org.opencontainers.image.version = "1.0.0"
@ -13,24 +13,23 @@ RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/

 RUN apt-get update && apt-get install -y wget libxml2 cuda-minimal-build-11-3 libcusparse-dev-11-3 libcublas-dev-11-3 libcusolver-dev-11-3 git
 RUN wget -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
+    "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
 ENV PATH /opt/conda/bin:$PATH

 COPY environment.yml /opt/openfold/environment.yml

 # installing into the base environment since the docker container wont do anything other than run openfold
-RUN conda env update -n base --file /opt/openfold/environment.yml && conda clean --all
+RUN mamba env update -n base --file /opt/openfold/environment.yml && mamba clean --all
+RUN export LD_LIBRARY_PATH=${CONDA_PREFIX}/lib:${LD_LIBRARY_PATH}

 COPY openfold /opt/openfold/openfold
 COPY scripts /opt/openfold/scripts
 COPY run_pretrained_openfold.py /opt/openfold/run_pretrained_openfold.py
 COPY train_openfold.py /opt/openfold/train_openfold.py
 COPY setup.py /opt/openfold/setup.py
-COPY lib/openmm.patch /opt/openfold/lib/openmm.patch
 RUN wget -q -P /opt/openfold/openfold/resources \
    https://git.scicore.unibas.ch/schwede/openstructure/-/raw/7102c63615b64735c4941278d92b554ec94415f8/modules/mol/alg/src/stereo_chemical_props.txt
-RUN patch -p0 -d /opt/conda/lib/python3.9/site-packages/ < /opt/openfold/lib/openmm.patch
 WORKDIR /opt/openfold
 RUN python3 setup.py install
--- a/README.md
+++ b/README.md
@ -48,37 +48,19 @@ and one of {`jackhmmer`, [MMseqs2](https://github.com/soedinglab/mmseqs2) (night
 installed on on your system. You'll need `git-lfs` to download OpenFold parameters. 
 Finally, some download scripts require `aria2c` and `aws`.

-For convenience, we provide a script that installs Miniconda locally, creates a 
-`conda` virtual environment, installs all Python dependencies, and downloads
-useful resources, including both sets of model parameters. Run:
+This package is currently supported for CUDA 11 and Pytorch 1.12

-```bash
-scripts/install_third_party_dependencies.sh
-```
+To install:
+1. Clone the repository, e.g. `git clone https://github.com/aqlaboratory/openfold.git`
+1. From the `openfold` repo: 
+    - Create a [Mamba]("https://github.com/conda-forge/miniforge/releases/latest/download/) environment, e.g. 
+        `mamba env create -n openfold_env -f environment.yml`
+      Mamba is recommended as the dependencies required by OpenFold are quite large and mamba can speed up the process.
+    - Activate the environment, e.g `conda activate openfold_env` 
+1. Run `scripts/install_third_party_dependencies.sh` to configure kernels and folding resources.

-To activate the environment, run:
+For some systems, it may help to append the Conda environment library path to `$LD_LIBRARY_PATH`. The `install_third_party_dependencies.sh` script does this once, but you may need this for each bash instance.

-```bash
-source scripts/activate_conda_env.sh
-```
-
-To deactivate it, run:
-
-```bash
-source scripts/deactivate_conda_env.sh
-```
-
-With the environment active, compile OpenFold's CUDA kernels with
-
-```bash
-python3 setup.py install
-```
-
-To install the HH-suite to `/usr/bin`, run
-
-```bash
-# scripts/install_hh_suite.sh
-```

 ## Usage

--- a/environment.yml
+++ b/environment.yml
@ -1,32 +1,37 @@
-name: openfold_venv
+name: openfold-venv
 channels:
  - conda-forge
  - bioconda
  - pytorch
 dependencies:
-  - conda-forge::python=3.9
-  - conda-forge::setuptools=59.5.0
-  - conda-forge::pip
-  - conda-forge::openmm=7.5.1
-  - conda-forge::pdbfixer
-  - conda-forge::cudatoolkit==11.3.*
+  - python=3.9
+  - libgcc=7.2
+  - setuptools=59.5.0
+  - pip
+  - openmm=7.7
+  - pdbfixer
+  - cudatoolkit==11.3.*
+  - pytorch-lightning==1.5.10
+  - biopython==1.79
+  - pandas==2.0.2
+  - numpy==1.21
+  - PyYAML==5.4.1
+  - requests
+  - scipy==1.7
+  - tqdm==4.62.2
+  - typing-extensions==3.10
+  - wandb==0.12.21
+  - modelcif==0.7
+  - awscli
+  - ml-collections
+  - aria2
+  - git
  - bioconda::hmmer==3.3.2
  - bioconda::hhsuite==3.3.0
  - bioconda::kalign2==2.04
  - pytorch::pytorch=1.12.*
  - pip:
-      - biopython==1.79
      - deepspeed==0.5.10
      - dm-tree==0.1.6
-      - ml-collections==0.1.0
-      - pandas==2.0.2
-      - numpy==1.21.2
-      - PyYAML==5.4.1
-      - requests==2.26.0
-      - scipy==1.7.1
-      - tqdm==4.62.2
-      - typing-extensions==3.10.0.2
-      - pytorch_lightning==1.5.10
-      - wandb==0.12.21
-      - modelcif==0.7
      - git+https://github.com/NVIDIA/dllogger.git
+      - git+https://github.com/Dao-AILab/flash-attention.git@5b838a8
--- a/lib/openmm.patch
+++ b/lib/openmm.patch
@ -1,42 +0,0 @@
-Index: simtk/openmm/app/topology.py
-===================================================================
--- simtk.orig/openmm/app/topology.py
-+++ simtk/openmm/app/topology.py
-@@ -356,19 +356,35 @@
-         def isCyx(res):
-             names = [atom.name for atom in res._atoms]
-             return 'SG' in names and 'HG' not in names
-+        # This function is used to prevent multiple di-sulfide bonds from being
-+        # assigned to a given atom. This is a DeepMind modification.
-+        def isDisulfideBonded(atom):
-+            for b in self._bonds:
-+                if (atom in b and b[0].name == 'SG' and
-+                    b[1].name == 'SG'):
-+                    return True
-+
-+            return False
- 
-         cyx = [res for res in self.residues() if res.name == 'CYS' and isCyx(res)]
-         atomNames = [[atom.name for atom in res._atoms] for res in cyx]
-         for i in range(len(cyx)):
-             sg1 = cyx[i]._atoms[atomNames[i].index('SG')]
-             pos1 = positions[sg1.index]
-+            candidate_distance, candidate_atom = 0.3*nanometers, None
-             for j in range(i):
-                 sg2 = cyx[j]._atoms[atomNames[j].index('SG')]
-                 pos2 = positions[sg2.index]
-                 delta = [x-y for (x,y) in zip(pos1, pos2)]
-                 distance = sqrt(delta[0]*delta[0] + delta[1]*delta[1] + delta[2]*delta[2])
-                if distance < 0.3*nanometers:
-                    self.addBond(sg1, sg2)
-+                if distance < candidate_distance and not isDisulfideBonded(sg2):
-+                    candidate_distance = distance
-+                    candidate_atom = sg2
-+            # Assign bond to closest pair.
-+            if candidate_atom:
-+                self.addBond(sg1, candidate_atom)
-+
-+
- 
- class Chain(object):
-     """A Chain object represents a chain within a Topology."""
--- a/openfold/np/relax/amber_minimize.py
+++ b/openfold/np/relax/amber_minimize.py
@ -28,18 +28,10 @@ import openfold.utils.loss as loss
 from openfold.np.relax import cleanup, utils
 import ml_collections
 import numpy as np
-try:
-    # openmm >= 7.6
-    import openmm
-    from openmm import unit
-    from openmm import app as openmm_app
-    from openmm.app.internal.pdbstructure import PdbStructure
-except ImportError:
-    # openmm < 7.6 (requires DeepMind patch)
-    from simtk import openmm
-    from simtk import unit
-    from simtk.openmm import app as openmm_app
-    from simtk.openmm.app.internal.pdbstructure import PdbStructure
+import openmm
+from openmm import unit
+from openmm import app as openmm_app
+from openmm.app.internal.pdbstructure import PdbStructure

 ENERGY = unit.kilocalories_per_mole
 LENGTH = unit.angstroms
--- a/openfold/np/relax/cleanup.py
+++ b/openfold/np/relax/cleanup.py
@ -20,14 +20,8 @@ cases like removing chains of length one (see clean_structure).
 import io

 import pdbfixer
-try:
-    # openmm >= 7.6
-    from openmm import app
-    from openmm.app import element
-except ImportError:
-    # openmm < 7.6 (requires DeepMind patch)
-    from simtk.openmm import app
-    from simtk.openmm.app import element
+from openmm import app
+from openmm.app import element


 def fix_pdb(pdbfile, alterations_info):
--- a/openfold/np/relax/utils.py
+++ b/openfold/np/relax/utils.py
@ -18,14 +18,8 @@ import io
 from openfold.np import residue_constants
 from Bio import PDB
 import numpy as np
-try:
-    # openmm >= 7.6
-    from openmm import app as openmm_app
-    from openmm.app.internal.pdbstructure import PdbStructure
-except ImportError:
-    # openmm < 7.6 (requires DeepMind patch)
-    from simtk.openmm import app as openmm_app
-    from simtk.openmm.app.internal.pdbstructure import PdbStructure
+from openmm import app as openmm_app
+from openmm.app.internal.pdbstructure import PdbStructure


 def overwrite_pdb_coordinates(pdb_str: str, pos) -> str:
--- a/scripts/install_third_party_dependencies.sh
+++ b/scripts/install_third_party_dependencies.sh
@ -1,49 +1,18 @@
 #!/bin/bash
-CONDA_INSTALL_URL=${CONDA_INSTALL_URL:-"https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh"}
-
-source scripts/vars.sh
-
-# Install Miniconda locally
-rm -rf lib/conda
-rm -f /tmp/Miniconda3-latest-Linux-x86_64.sh
-wget -P /tmp \
-    "${CONDA_INSTALL_URL}" \
-    && bash /tmp/Miniconda3-latest-Linux-x86_64.sh -b -p lib/conda \
-    && rm /tmp/Miniconda3-latest-Linux-x86_64.sh
-
-# Grab conda-only packages
-export PATH=lib/conda/bin:$PATH
-lib/conda/bin/python3 -m pip install nvidia-pyindex
-conda env create --name=${ENV_NAME} -f environment.yml
-source scripts/activate_conda_env.sh
-
-echo "Attempting to install FlashAttention"
-git clone https://github.com/HazyResearch/flash-attention
-CUR_DIR=$PWD
-cd flash-attention
-git checkout 5b838a8bef
-python3 setup.py install
-cd $CUR_DIR
-
-# Install DeepMind's OpenMM patch
-OPENFOLD_DIR=$PWD
-pushd lib/conda/envs/$ENV_NAME/lib/python3.9/site-packages/ \
-    && patch -p0 < $OPENFOLD_DIR/lib/openmm.patch \
-    && popd

 # Download folding resources
-wget --no-check-certificate -P openfold/resources \
+wget -N --no-check-certificate -P openfold/resources \
    https://git.scicore.unibas.ch/schwede/openstructure/-/raw/7102c63615b64735c4941278d92b554ec94415f8/modules/mol/alg/src/stereo_chemical_props.txt

 # Certain tests need access to this file
 mkdir -p tests/test_data/alphafold/common
 ln -rs openfold/resources/stereo_chemical_props.txt tests/test_data/alphafold/common

-echo "Downloading OpenFold parameters..."
-bash scripts/download_openfold_params.sh openfold/resources
-
-echo "Downloading AlphaFold parameters..."
-bash scripts/download_alphafold_params.sh openfold/resources
-
 # Decompress test data
 gunzip -c tests/test_data/sample_feats.pickle.gz > tests/test_data/sample_feats.pickle
+
+python setup.py install
+
+export LD_LIBRARY_PATH=$CONDA_PREFIX/lib:$LD_LIBRARY_PATH
+# This setting is used to fix a worker assignment issue during data loading
+conda env config vars set KMP_AFFINITY=none