easybuilders · zemu-unile · Aug 23, 2021 · Aug 24, 2021 · Aug 27, 2021 · Aug 27, 2021
diff --git a/...eTTAFold/RoseTTAFold-1.0.0-fosscuda-2019b-Python-3.7.4-PyTorch-1.8.1-TensorFlow-1.15.2.eb b/...eTTAFold/RoseTTAFold-1.0.0-fosscuda-2019b-Python-3.7.4-PyTorch-1.8.1-TensorFlow-1.15.2.eb
@@ -0,0 +1,93 @@
+easyblock = 'Tarball'
+
+name = 'RoseTTAFold'
+version = '1.0.0'
+local_pytorch_version = '1.8.1'
+local_pytorch_suffix = '-PyTorch-%s' % local_pytorch_version
+local_python_suffix = '-Python-%(pyver)s'
+local_tf_version = '1.15.2'
+local_tf_suffix = '-TensorFlow-%s' % local_tf_version
+versionsuffix = local_python_suffix + local_pytorch_suffix + local_tf_suffix
+
+homepage = 'https://github.com/RosettaCommons/RoseTTAFold'
+description = """Official implementation of RoseTTAFold: Accurate prediction
+of protein structures and interactions using a 3-track network."""
+
+toolchain = {'name': 'fosscuda', 'version': '2019b'}
+toolchainopts = {'openmp': True}
+
+github_account = 'RosettaCommons'
+source_urls = [GITHUB_SOURCE]
+
+# See https://github.com/RosettaCommons/RoseTTAFold/#installation for downloading weights.tar.gz
+#
+# additionally needs some protein databases to work, total size about 2.5 TB (~500 GB with file system compression)
+# See https://github.com/RosettaCommons/RoseTTAFold/blob/main/README.md for download links, as of August 2021:
+# http://wwwuser.gwdg.de/~compbiol/uniclust/2020_06/UniRef30_2020_06_hhsuite.tar.gz
+# https://bfd.mmseqs.com/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt.tar.gz
+# https://files.ipd.uw.edu/pub/RoseTTAFold/pdb100_2021Mar03.tar.gz
+sources = [
+    'v%(version)s.tar.gz',
+    'weights.tar.gz',
+]
+
+patches = [
+    'RoseTTAFold-1.0.0_cpu_mem_from_env.patch',
+    'RoseTTAFold-1.0.0_db_paths_from_env.patch',
+    'RoseTTAFold-1.0.0_no_conda.patch',
+    'RoseTTAFold-1.0.0_use_eb_paths.patch',
+    'RoseTTAFold-1.0.0_fix_cache_directory.patch',
+    'RoseTTAFold-1.0.0_lddt_path.patch',
+]
+
+checksums = [
+    'c3db4ec7a0d933c5877a76112d51aa0466d86698807603614c419286261df524',  # v1.0.0.tar.gz
+    '6379adffe1e0a26fcaa33a228f695be9a10deeba565b38cb0b1b9231d8b5c369',  # weights.tar.gz
+    '06de539538fb48ce2efdf3a989f50f02fd360556b76df18f3e0083af5d573461',  # RoseTTAFold-1.0.0_cpu_mem_from_env.patch
+    'f7a163cdf7d2c62e18a691a322c81c2fd20a2f92bdbd6e806f93246842092e3e',  # RoseTTAFold-1.0.0_db_paths_from_env.patch
+    'a1ea483525cd94712e56107de6513c0c27be8942fd9f3ad627482d787b3c3a50',  # RoseTTAFold-1.0.0_no_conda.patch
+    '9d4d56993e5e1bc52ecb3c67b3f71e3b869d559c0c85281750bc0a708b271d55',  # RoseTTAFold-1.0.0_use_eb_paths.patch
+    '6e7abdce5c677803aeb174b527d1a360212ef4cc494cf949f75f6baeaffc29b5',  # RoseTTAFold-1.0.0_fix_cache_directory.patch
+    '39c56e08b27b2a570d0c920cf1150326ddd583d467ae530a6ba559c183f110d6',  # RoseTTAFold-1.0.0_lddt_path.patch
+]
+
+dependencies = [
+    ('Biopython', '1.75', local_python_suffix),
+    ('BLAST', '2.2.26', '-Linux_x86_64', SYSTEM),
+    ('HH-suite', '3.3.0', local_python_suffix),
+    ('lDDT', '1.2', '', SYSTEM),
+    ('PyRosetta', '4.release-292', local_python_suffix),
+    ('PyTorch-Geometric', '1.6.3', local_python_suffix + local_pytorch_suffix),
+    ('Python', '3.7.4'),
+    ('SciPy-bundle', '2019.10', local_python_suffix),
+    ('TensorFlow', local_tf_version, local_python_suffix),
+    ('CSBLAST', '2.2.3'),
+    ('DGL', '0.6.1', local_python_suffix + local_pytorch_suffix),
+    ('lie_learn', '0.0.1.post1', local_python_suffix),
+    ('parallel', '20190922'),
+    ('PSIPRED', '4.02'),
+]
+
+local_broken_symlinks = ['example/pyrosetta/model/model_%s.pdb' % i for i in range(1, 6)]
+preinstallopts = 'cp -r ../weights . && rm %s && ' % ' '.join(local_broken_symlinks)
+
+modextrapaths = {'PATH': ''}
+
+sanity_check_paths = {
+    'dirs': [
+        'DAN-msa',
+        'folding',
+        'network',
+        'weights',
+    ],
+    'files': [
+        'run_e2e_ver.sh',
+        'run_pyrosetta_ver.sh',
+        'input_prep/make_ss.sh',
+        'input_prep/make_msa.sh',
+        'network/predict_e2e.py',
+        'network/predict_pyRosetta.py',
+    ]
+}
+
+moduleclass = 'bio'
diff --git a/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_cpu_mem_from_env.patch b/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_cpu_mem_from_env.patch
@@ -0,0 +1,31 @@
+Read the number of cores and maximum memory to use from environment instead of hard coding
+author: Christoph Siegert (Leipzig University)
+
+diff -ruN RoseTTAFold.orig/run_e2e_ver.sh RoseTTAFold/run_e2e_ver.sh
+--- RoseTTAFold.orig/run_e2e_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_e2e_ver.sh	2021-08-16 13:06:25.547251482 +0200
+@@ -15,8 +15,8 @@
+ SCRIPT=`realpath -s $0`
+ export PIPEDIR=`dirname $SCRIPT`
+
+-CPU="8"  # number of CPUs to use
+-MEM="64" # max memory (in GB)
++CPU=${CPU:-"8"}  # number of CPUs to use
++MEM=${MEM:-"64"} # max memory (in GB)
+
+ # Inputs:
+ IN="$1"                # input.fasta
+diff -ruN RoseTTAFold.orig/run_pyrosetta_ver.sh RoseTTAFold/run_pyrosetta_ver.sh
+--- RoseTTAFold.orig/run_pyrosetta_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_pyrosetta_ver.sh	2021-08-16 13:07:02.619152279 +0200
+@@ -15,8 +15,8 @@
+ SCRIPT=`realpath -s $0`
+ export PIPEDIR=`dirname $SCRIPT`
+
+-CPU="8"  # number of CPUs to use
+-MEM="64" # max memory (in GB)
++CPU=${CPU:-"8"}  # number of CPUs to use
++MEM=${MEM:-"64"} # max memory (in GB)
+
+ # Inputs:
+ IN="$1"                # input.fasta
diff --git a/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_db_paths_from_env.patch b/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_db_paths_from_env.patch
@@ -0,0 +1,61 @@
+Read the database path from environment instead of assuming them in the installation directory
+author: Christoph Siegert (Leipzig University)
+
+diff -ruN RoseTTAFold.orig/input_prep/make_msa.sh RoseTTAFold/input_prep/make_msa.sh
+--- RoseTTAFold.orig/input_prep/make_msa.sh	2021-08-16 12:54:05.769231102 +0200
++++ RoseTTAFold/input_prep/make_msa.sh	2021-08-16 13:22:02.508663769 +0200
+@@ -9,8 +9,8 @@
+ MEM="$4"
+
+ # sequence databases
+-DB="$PIPEDIR/UniRef30_2020_06/UniRef30_2020_06"
+-MYDB="$PIPEDIR/bfd/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt"
++DB="$UNIREF30PATH/UniRef30_2020_06"
++MYDB="$BFDPATH/bfd_metaclust_clu_complete_id30_c90_final_seq.sorted_opt"
+
+ # setup hhblits command
+ HHBLITS="hhblits -o /dev/null -mact 0.35 -maxfilt 100000000 -neffmax 20 -cov 25 -cpu $CPU -nodiff -realign_max 100000000 -maxseq 1000000 -maxmem $MEM -n 4 -d $DB -d $MYDB"
+diff -ruN RoseTTAFold.orig/run_e2e_ver.sh RoseTTAFold/run_e2e_ver.sh
+--- RoseTTAFold.orig/run_e2e_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_e2e_ver.sh	2021-08-16 13:21:02.188832795 +0200
+@@ -22,6 +22,9 @@
+ IN="$1"                # input.fasta
+ WDIR=`realpath -s $2`  # working folder
+
++BFDPATH=${BFDPATH:-"$PIPEDIR/bfd/"}
++PDB100PATH=${PDB100PATH:-"$PIPEDIR/pdb100_2021Mar03/"}
++UNIREF30PATH=${UNIREF30PATH:-"$PIPEDIR/UniRef30_2020_06/"}
+
+ LEN=`tail -n1 $IN | wc -m`
+
+@@ -51,7 +54,7 @@
+ ############################################################
+ # 3. search for templates
+ ############################################################
+-DB="$PIPEDIR/pdb100_2021Mar03/pdb100_2021Mar03"
++DB="$PDB100PATH/pdb100_2021Mar03"
+ if [ ! -s $WDIR/t000_.hhr ]
+ then
+     echo "Running hhsearch"
+diff -ruN RoseTTAFold.orig/run_pyrosetta_ver.sh RoseTTAFold/run_pyrosetta_ver.sh
+--- RoseTTAFold.orig/run_pyrosetta_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_pyrosetta_ver.sh	2021-08-16 13:21:37.380734181 +0200
+@@ -22,6 +22,9 @@
+ IN="$1"                # input.fasta
+ WDIR=`realpath -s $2`  # working folder
+
++BFDPATH=${BFDPATH:-"$PIPEDIR/bfd/"}
++PDB100PATH=${PDB100PATH:-"$PIPEDIR/pdb100_2021Mar03/"}
++UNIREF30PATH=${UNIREF30PATH:-"$PIPEDIR/UniRef30_2020_06/"}
+
+ LEN=`tail -n1 $IN | wc -m`
+
+@@ -51,7 +54,7 @@
+ ############################################################
+ # 3. search for templates
+ ############################################################
+-DB="$PIPEDIR/pdb100_2021Mar03/pdb100_2021Mar03"
++DB="$PDB100PATH/pdb100_2021Mar03"
+ if [ ! -s $WDIR/t000_.hhr ]
+ then
+     echo "Running hhsearch"
diff --git a/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_fix_cache_directory.patch b/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_fix_cache_directory.patch
@@ -0,0 +1,16 @@
+Changes the cache directory path from the installation directory to the users home directory
+to avoid a crash due to insufficient permissions.
+author: Christoph Siegert (Leipzig University)
+
+diff -ruN RoseTTAFold.orig/network/equivariant_attention/from_se3cnn/utils_steerable.py RoseTTAFold/network/equivariant_attention/from_se3cnn/utils_steerable.py
+--- RoseTTAFold.orig/network/equivariant_attention/from_se3cnn/utils_steerable.py	2021-08-16 12:54:05.769231102 +0200
++++ RoseTTAFold/network/equivariant_attention/from_se3cnn/utils_steerable.py	2021-08-18 16:17:10.653216846 +0200
+@@ -33,7 +33,7 @@
+     return get_matrix_kernel(torch.cat(As, dim=0), eps)
+
+
+-@cached_dirpklgz("%s/cache/trans_Q"%os.path.dirname(os.path.realpath(__file__)))
++@cached_dirpklgz("%s/.cache/trans_Q"%os.path.expanduser('~'))
+ def _basis_transformation_Q_J(J, order_in, order_out, version=3):  # pylint: disable=W0613
+     """
+     :param J: order of the spherical harmonics
diff --git a/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_lddt_path.patch b/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_lddt_path.patch
@@ -0,0 +1,27 @@
+Pick lddt excetutable from PATH instead of script directory
+author: Christoph Siegert (Leipzig University)
+
+diff --git a/DAN-msa/pick_final_models.div.py b/DAN-msa/pick_final_models.div.py
+index c1f2e9d..4600837 100644
+--- a/DAN-msa/pick_final_models.div.py
++++ b/DAN-msa/pick_final_models.div.py
+@@ -7,8 +7,6 @@ import numpy as np
+ from sklearn.cluster import AgglomerativeClustering
+ import multiprocessing as mp
+
+-script_dir = "/".join(os.path.dirname(os.path.realpath(__file__)).split('/')[:-1])
+-
+ def smooth(x, window_len=13, window='hanning'):
+     s = np.r_[[x[0]]*(window_len//2), x, [x[-1]]*(window_len//2)]
+     if window == 'flat': #moving average
+@@ -69,8 +67,8 @@ def calc_lddt_dist(args):
+     pose_i = pose_s[i]
+     pose_j = pose_s[j]
+     #
+-    lddt_1 = float(os.popen("%s/lddt/lddt -c %s %s | grep Glob"%(script_dir, pose_i, pose_j)).readlines()[-1].split()[-1])
+-    lddt_2 = float(os.popen("%s/lddt/lddt -c %s %s | grep Glob"%(script_dir, pose_j, pose_i)).readlines()[-1].split()[-1])
++    lddt_1 = float(os.popen("lddt -c %s %s | grep Glob"%(pose_i, pose_j)).readlines()[-1].split()[-1])
++    lddt_2 = float(os.popen("lddt -c %s %s | grep Glob"%(pose_j, pose_i)).readlines()[-1].split()[-1])
+     lddt = (lddt_1 + lddt_2) / 2.0
+     return 1 - lddt
+
diff --git a/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_no_conda.patch b/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_no_conda.patch
@@ -0,0 +1,67 @@
+Do not attempt to use the conda environment
+author: Christoph Siegert (Leipzig University)
+
+diff -ruN RoseTTAFold.orig/run_e2e_ver.sh RoseTTAFold/run_e2e_ver.sh
+--- RoseTTAFold.orig/run_e2e_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_e2e_ver.sh	2021-08-16 12:55:10.969056632 +0200
+@@ -3,15 +3,6 @@
+ # make the script stop when error (non-true exit code) is occured
+ set -e
+
+-############################################################
+-# >>> conda initialize >>>
+-# !! Contents within this block are managed by 'conda init' !!
+-__conda_setup="$('conda' 'shell.bash' 'hook' 2> /dev/null)"
+-eval "$__conda_setup"
+-unset __conda_setup
+-# <<< conda initialize <<<
+-############################################################
+-
+ SCRIPT=`realpath -s $0`
+ export PIPEDIR=`dirname $SCRIPT`
+
+@@ -27,7 +18,6 @@
+
+ mkdir -p $WDIR/log
+
+-conda activate RoseTTAFold
+ ############################################################
+ # 1. generate MSAs
+ ############################################################
+diff -ruN RoseTTAFold.orig/run_pyrosetta_ver.sh RoseTTAFold/run_pyrosetta_ver.sh
+--- RoseTTAFold.orig/run_pyrosetta_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_pyrosetta_ver.sh	2021-08-16 12:55:43.860968614 +0200
+@@ -3,15 +3,6 @@
+ # make the script stop when error (non-true exit code) is occured
+ set -e
+
+-############################################################
+-# >>> conda initialize >>>
+-# !! Contents within this block are managed by 'conda init' !!
+-__conda_setup="$('conda' 'shell.bash' 'hook' 2> /dev/null)"
+-eval "$__conda_setup"
+-unset __conda_setup
+-# <<< conda initialize <<<
+-############################################################
+-
+ SCRIPT=`realpath -s $0`
+ export PIPEDIR=`dirname $SCRIPT`
+
+@@ -27,7 +18,6 @@
+
+ mkdir -p $WDIR/log
+
+-conda activate RoseTTAFold
+ ############################################################
+ # 1. generate MSAs
+ ############################################################
+@@ -81,9 +71,6 @@
+ ############################################################
+ mkdir -p $WDIR/pdb-3track
+
+-conda deactivate
+-conda activate folding
+-
+ for m in 0 1 2
+ do
+     for p in 0.05 0.15 0.25 0.35 0.45
diff --git a/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_use_eb_paths.patch b/easybuild/easyconfigs/r/RoseTTAFold/RoseTTAFold-1.0.0_use_eb_paths.patch
@@ -0,0 +1,49 @@
+Use easy build installation roots
+author: Christoph Siegert (Leipzig University)
+
+diff -ruN RoseTTAFold.orig/input_prep/make_ss.sh RoseTTAFold/input_prep/make_ss.sh
+--- RoseTTAFold.orig/input_prep/make_ss.sh	2021-08-16 12:54:05.769231102 +0200
++++ RoseTTAFold/input_prep/make_ss.sh	2021-08-16 13:03:22.003742639 +0200
+@@ -1,6 +1,6 @@
+ #!/bin/bash
+
+-DATADIR="$CONDA_PREFIX/share/psipred_4.01/data"
++DATADIR="$EBROOTPSIPRED/data"
+ echo $DATADIR
+
+ i_a3m="$1"
+@@ -8,7 +8,7 @@
+
+ ID=$(basename $i_a3m .a3m).tmp
+
+-$PIPEDIR/csblast-2.2.3/bin/csbuild -i $i_a3m -I a3m -D $PIPEDIR/csblast-2.2.3/data/K4000.crf -o $ID.chk -O chk
++csbuild -i $i_a3m -I a3m -D $EBROOTCSBLAST/data/K4000.crf -o $ID.chk -O chk
+
+ head -n 2 $i_a3m > $ID.fasta
+ echo $ID.chk > $ID.pn
+diff -ruN RoseTTAFold.orig/run_e2e_ver.sh RoseTTAFold/run_e2e_ver.sh
+--- RoseTTAFold.orig/run_e2e_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_e2e_ver.sh	2021-08-16 13:01:01.476118686 +0200
+@@ -12,8 +12,7 @@
+ # <<< conda initialize <<<
+ ############################################################
+
+-SCRIPT=`realpath -s $0`
+-export PIPEDIR=`dirname $SCRIPT`
++export PIPEDIR=$EBROOTROSETTAFOLD
+
+ CPU="8"  # number of CPUs to use
+ MEM="64" # max memory (in GB)
+diff -ruN RoseTTAFold.orig/run_pyrosetta_ver.sh RoseTTAFold/run_pyrosetta_ver.sh
+--- RoseTTAFold.orig/run_pyrosetta_ver.sh	2021-08-16 12:54:04.081235616 +0200
++++ RoseTTAFold/run_pyrosetta_ver.sh	2021-08-16 13:00:45.388161737 +0200
+@@ -12,8 +12,7 @@
+ # <<< conda initialize <<<
+ ############################################################
+
+-SCRIPT=`realpath -s $0`
+-export PIPEDIR=`dirname $SCRIPT`
++export PIPEDIR=$EBROOTROSETTAFOLD
+
+ CPU="8"  # number of CPUs to use
+ MEM="64" # max memory (in GB)
diff --git a/test/easyconfigs/easyconfigs.py b/test/easyconfigs/easyconfigs.py
@@ -496,6 +496,9 @@ def check_dep_vars(self, gen, dep, dep_vars):
             # egl variant of glew is required by libwpe
             'glew': [('2.2.0; versionsuffix: -egl', [r'libwpe-1\.13\.3-GCCcore-11\.2\.0'])],
             'Geant4': [('11.0.1;', [r'GATE-9\.2-foss-2021b'])],
+            # ncbi-vdb v2.x require HDF5 v1.10.x (HISAT2, SKESA, shovill depend on ncbi-vdb)
+            # RoseTTAFold on Ivy Bridge needs HH-Suite-3.3.0 or newer
+            'HH-Suite': [('3.3.0;', ['RoseTTAFold-1.0.0-'])],
             # ncbi-vdb v2.x and v3.0.0 require HDF5 v1.10.x (HISAT2, SKESA, shovill depend on ncbi-vdb)
             'HDF5': [
                 (r'1\.10\.', [r'ncbi-vdb-2\.11\.', r'ncbi-vdb-3\.0\.0', r'HISAT2-2\.2\.', r'SKESA-2\.4\.',
@@ -521,7 +524,7 @@ def check_dep_vars(self, gen, dep, dep_vars):
             'LLVM': [
                 # numba 0.47.x requires LLVM 7.x or 8.x (see https://github.com/numba/llvmlite#compatibility)
                 (r'8\.', [r'numba-0\.47\.0-', r'librosa-0\.7\.2-', r'BirdNET-20201214-',
-                          r'scVelo-0\.1\.24-', r'PyTorch-Geometric-1\.[346]\.[23]']),
+                          r'scVelo-0\.1\.24-', r'PyTorch-Geometric-1\.[346]\.[23]', 'RoseTTAFold-1.0.0-']),
                 (r'10\.0\.1', [r'cell2location-0\.05-alpha-', r'cryoDRGN-0\.3\.2-', r'loompy-3\.0\.6-',
                                r'numba-0\.52\.0-', r'PyOD-0\.8\.7-', r'PyTorch-Geometric-1\.6\.3',
                                r'scanpy-1\.7\.2-', r'umap-learn-0\.4\.6-']),