Skip to main content
Pre-release

This release is a pre-release and may not be stable for production use.

HPSandbox

HPSandbox is a set of Python objects that allow you to quickly write simple python scripts to explore the two-dimensional HP lattice model of proteins of Chan and Dill.

Last Updated: March 2012

GETTING STARTED

Installation

Please define the HPSANDBOXHOME environment variable to be this directory

In bash: $ export HPSANDBOXHOME=/Users/vince/scripts/HPSandbox/trunk/HPSandbox

COPYRIGHT

This python package is Copyright (C) 2007 Vincent Voelz voelz@temple.edu. Feel free to modify this code as needed, as long as you can keep it publicly available!

PACKAGE CONTENTS

  • Chain.py An object to represent the 2D HP lattice chain and its attributes, with method functions.
  • Config.py A data structure to hold configuration parameters.
  • Monty.py A collection of functions to perform Monte Carlo move-set operations on the Chain() object.
  • Replica.py A container object, to hold the Chain() and Monty() objects
  • Trajectory.py A set of functions for creating, reading, writing, and organizing trajectory files

/examples A directory of example scripts /sequences Containing descriptions of the native states of foldable sequences: /clist - contact state lists for chain lengths 10 through 21 /conf - coordinates (conformations) for chain lengths 10 through 19
COUNTS - text file counts of all unique (nonsymmetric) conformations for a given chain length

This package has been tested with Python 2.3 and 2.4. Older/newer versions may work too, but haven't been tested.

SETUP

In order to get these example scripts to work correctly, you need to set up the following:

  • The HPSandbox directory (i.e python module) must be defined in your PYTHONPATH environment variable
  • In the /examples folder mcrex.conf file needs to be changed to reflect the absolute pathname of the sequences/clist/hp** directory.

EXAMPLE SCRIPTS

Please see the /examples directory and the README file therein for some test scripts and examples showing how to use the HPSandbox function.

DOCUMENTATION

Documentation can be obtained using the pydoc standard module of python. For example:

from HPSandbox import *
import pydoc
pydoc.help(Chain)

Frequently Asked Questions (FAQ)

What can HPSandbox do?

HPSandbox can either 1) enumerate, or 2) perform Monte Carlo "dynamics" for 2-dimensional, square-lattice "bead-on-a-string" type chains.

How long a chain can I simulate?

It depends on how long you are willing to wait. For instance, all conformations of 16-mers can be enumerated in a few minutes on a typical personal computer. Each increase in chain length adds a factor of about 2.7 to the calculation.

Can I use other potentials besides the HP model?

Sure! But you'll have to put it in yourself. The code is designed for HP sequences, so if you want to study a model using beads of only two flavors, it is easy to just modify theMonty.energy()class function. More complicated models would require a more thorough, but straightforward reworking of the code.

What are the reference(s) for the HP 2D Model?

Lau, K.F. and K.A. Dill. A Lattice Statistical Mechanics Model of the Conformational and Sequence Spaces of Proteins. Macromolecules 22: 3986-3997 (1989). Dill, K.A., S. Bromberg, K. Yue, K.M. Fiebig, D.P. Yee, P.D. Thomas, and H.S. Chan. Principles of Protein Folding - A Perspective From Simple Exact Models. Protein Science 4: 561-602, 1995. Lau and Dill (1989) is the first use of the HP model, while Dill et al. (1995) is a more comprehensive review.

What movesets are used for the Monte Carlo routines in HPSandbox?

The movesets are described in Dill et al. Protein Science 4: 561-602, 1995:

Vincent Voelz voelz@temple.edu

Metadata

Release files for hpsandbox 1.0.0b0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for hpsandbox 1.0.0b0
File Size Uploaded Provenance Actions
hpsandbox-1.0.0b0.tar.gz 14.6 kB Without provenance Details for hpsandbox-1.0.0b0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for hpsandbox 1.0.0b0
File Interpreter ABI Platform Actions
hpsandbox-1.0.0b0-py3-none-any.whl Python 3 none any Details for hpsandbox-1.0.0b0-py3-none-any.whl

Total release size: 28.0 kB

Release files / hpsandbox-1.0.0b0.tar.gz

Download URL hpsandbox-1.0.0b0.tar.gz
Size 14.6 kB
Tags Source
SHA-256 checksum
How to use checksums
aa42664bcfa54c0feda1a2568a05e0c400edf7497450aa2af8f22ad539245b85
BLAKE2b-256 checksum
How to use checksums
4bb581f1370d4a98019167f70959873f296c777fc2b4b6738c8db3df25935169
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/4.0.2 CPython/3.8.13

Release files / hpsandbox-1.0.0b0-py3-none-any.whl

Download URL hpsandbox-1.0.0b0-py3-none-any.whl
Size 13.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
2bb3ea83dfed3eb74c2fb16849a23790bf3b1893b0a91c01fd886b63e8da02c4
BLAKE2b-256 checksum
How to use checksums
97b4296a96afcc1e7c9754d85aa6ad0162ec49b2d8c186c805cbb6ff254182fc
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/4.0.2 CPython/3.8.13

Security for hpsandbox 1.0.0b0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release
Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page