annotate scripts/download_pdb_mmcif.sh @ 0:7ae9d78b06f5 draft

"planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
author galaxy-australia
date Fri, 28 Jan 2022 04:56:29 +0000
parents
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
1 #!/bin/bash
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
2 #
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
3 # Copyright 2021 DeepMind Technologies Limited
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
4 #
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
5 # Licensed under the Apache License, Version 2.0 (the "License");
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
6 # you may not use this file except in compliance with the License.
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
7 # You may obtain a copy of the License at
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
8 #
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
9 # http://www.apache.org/licenses/LICENSE-2.0
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
10 #
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
11 # Unless required by applicable law or agreed to in writing, software
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
12 # distributed under the License is distributed on an "AS IS" BASIS,
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
13 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
14 # See the License for the specific language governing permissions and
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
15 # limitations under the License.
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
16 #
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
17 # Downloads, unzips and flattens the PDB database for AlphaFold.
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
18 #
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
19 # Usage: bash download_pdb_mmcif.sh /path/to/download/directory
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
20 set -e
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
21
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
22 if [[ $# -eq 0 ]]; then
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
23 echo "Error: download directory must be provided as an input argument."
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
24 exit 1
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
25 fi
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
26
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
27 if ! command -v aria2c &> /dev/null ; then
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
28 echo "Error: aria2c could not be found. Please install aria2c (sudo apt install aria2)."
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
29 exit 1
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
30 fi
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
31
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
32 if ! command -v rsync &> /dev/null ; then
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
33 echo "Error: rsync could not be found. Please install rsync."
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
34 exit 1
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
35 fi
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
36
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
37 DOWNLOAD_DIR="$1"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
38 ROOT_DIR="${DOWNLOAD_DIR}/pdb_mmcif"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
39 RAW_DIR="${ROOT_DIR}/raw"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
40 MMCIF_DIR="${ROOT_DIR}/mmcif_files"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
41
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
42 echo "Running rsync to fetch all mmCIF files (note that the rsync progress estimate might be inaccurate)..."
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
43 mkdir --parents "${RAW_DIR}"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
44 rsync --recursive --links --perms --times --compress --info=progress2 --delete --port=33444 \
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
45 rsync.rcsb.org::ftp_data/structures/divided/mmCIF/ \
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
46 "${RAW_DIR}"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
47
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
48 echo "Unzipping all mmCIF files..."
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
49 find "${RAW_DIR}/" -type f -iname "*.gz" -exec gunzip {} +
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
50
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
51 echo "Flattening all mmCIF files..."
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
52 mkdir --parents "${MMCIF_DIR}"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
53 find "${RAW_DIR}" -type d -empty -delete # Delete empty directories.
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
54 for subdir in "${RAW_DIR}"/*; do
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
55 mv "${subdir}/"*.cif "${MMCIF_DIR}"
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
56 done
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
57
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
58 # Delete empty download directory structure.
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
59 find "${RAW_DIR}" -type d -empty -delete
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
60
7ae9d78b06f5 "planemo upload for repository https://github.com/usegalaxy-au/galaxy-local-tools commit 7b79778448363aa8c9b14604337e81009e461bd2-dirty"
galaxy-australia
parents:
diff changeset
61 aria2c "ftp://ftp.wwpdb.org/pub/pdb/data/status/obsolete.dat" --dir="${ROOT_DIR}"