PDB Datasets

In short

This tutorial shows how to create datasets with PDBs through the .zarr format.

This feature is still very new.

The features we will show in this tutorial are still experimental. We would love to learn from the community how we can make it easier to create datasets.

Dummy PDB example¶

In [1]:

import platformdirs

import datamol as dm

from polaris.dataset import DatasetFactory
from polaris.dataset.converters import PDBConverter

SAVE_DIR = dm.fs.join(platformdirs.user_cache_dir(appname="polaris-tutorials"), "dataset_pdb")

import platformdirs import datamol as dm from polaris.dataset import DatasetFactory from polaris.dataset.converters import PDBConverter SAVE_DIR = dm.fs.join(platformdirs.user_cache_dir(appname="polaris-tutorials"), "dataset_pdb")

Fetch PDB files from RCSB PDB¶

In [ ]:

import biotite.database.rcsb as rcsb

pdb_path = rcsb.fetch("6s89", "pdb", SAVE_DIR)
print(pdb_path)

import biotite.database.rcsb as rcsb pdb_path = rcsb.fetch("6s89", "pdb", SAVE_DIR) print(pdb_path)

Create dataset from PDB file¶

In [14]:

save_dst = dm.fs.join(SAVE_DIR, "tutorial_pdb.zarr")

factory = DatasetFactory(zarr_root_path=save_dst)
factory.reset(save_dst)

factory.register_converter("pdb", PDBConverter(pdb_column="pdb"))
factory.add_from_file(pdb_path)

# Build the dataset
dataset = factory.build()

save_dst = dm.fs.join(SAVE_DIR, "tutorial_pdb.zarr") factory = DatasetFactory(zarr_root_path=save_dst) factory.reset(save_dst) factory.register_converter("pdb", PDBConverter(pdb_column="pdb")) factory.add_from_file(pdb_path) # Build the dataset dataset = factory.build()

Check the dataset¶

In [15]:

dataset

dataset

Out[15]:

name	None
description
tags
user_attributes
owner	None
polaris_version	0.7.10.dev22+g8edf177.d20240814
default_adapters	pdb ARRAY_TO_PDB
zarr_root_path	/Users/lu.zhu/Library/Caches/polaris-tutorials/002/tutorial_pdb.zarr
readme
annotations	pdb is_pointer True modality PROTEIN_3D description None user_attributes dtype object
source	None
license	None
curation_reference	None
cache_dir	/Users/lu.zhu/Library/Caches/polaris/datasets/b0895f92-5a11-4e48-953f-3f969c6a9ca6
md5sum	66f3c7774e655bc6d48c907100d6912f
artifact_id	None
n_rows	1
n_columns	1

Check data table¶

In [16]:

dataset.table

dataset.table

Out[16]:

	pdb
0	pdb/6s89

Get PDB data from specific row¶

A array of list of biotite.Atom will be returned. See more details at fastpdb and Atom.

In [ ]:

dataset.get_data(0, "pdb")

dataset.get_data(0, "pdb")

Create dataset from multiple PDB files¶

In [7]:

pdb_paths = rcsb.fetch(["1l2y", "4i23"], "pdb", SAVE_DIR)
print(pdb_paths)

pdb_paths = rcsb.fetch(["1l2y", "4i23"], "pdb", SAVE_DIR) print(pdb_paths)

['/Users/lu.zhu/Library/Caches/polaris-tutorials/002/1l2y.pdb', '/Users/lu.zhu/Library/Caches/polaris-tutorials/002/4i23.pdb']

In [8]:

factory = DatasetFactory(SAVE_DIR.join("pdbs.zarr"))

converter = PDBConverter()
factory.register_converter("pdb", converter)

factory.add_from_files(pdb_paths, axis=0)
dataset = factory.build()

factory = DatasetFactory(SAVE_DIR.join("pdbs.zarr")) converter = PDBConverter() factory.register_converter("pdb", converter) factory.add_from_files(pdb_paths, axis=0) dataset = factory.build()

In [9]:

dataset.table

dataset.table

Out[9]:

	pdb
0	pdb/1l2y
1	pdb/4i23

In [ ]:

dataset.get_data(1, "pdb")

dataset.get_data(1, "pdb")

The process of completing the dataset's metadata and uploading it to the hub follows the same steps as outlined in the tutorial dataset_zarr.ipynb

The End.