txt2contincd
txt2contincd convert raw text data files into a CONTIN-CD input file. It use maidenhair for reading raw text files so any kind of raw text file can be used if there is a maidenhair plugins.
Installation
Use pip like:
$ pip install txt2contincd
Quick Usage
Assume that you have measured the far-UV CD spectrum with the following condition:
The number of residues (amino acids): 260 aa The molecular weight of the protein: 29.07 kDa The concentration of the protein: 0.303 mg/mL The length of the light pathway: 0.1 cm
Then run txt2contincd with
% txt2contincd -n 260 -m 29.07 -c 0.303 -L 0.1 <raw CD spectrum>
It will produce contin-cd.in file.
Usage
usage: txt2contincd [-h] [-v] [-p PARSER] [-l LOADER] [-u USING] [-a] [-s]
[-o OUTPUT] [-n NUMBER] [-m MOLECULAR_WEIGHT]
[-c CONCENTRATION]
[--molar-concentration MOLAR_CONCENTRATION] [-L LENGTH]
pathname
positional arguments:
pathname An unix grob style filename pattern for the data files
optional arguments:
-h, --help show this help message and exit
-v, --version show program's version number and exit
Reading options:
-p PARSER, --parser PARSER
A maidenhair parser name which will be used to parse
the raw text data.
-l LOADER, --loader LOADER
A maidenhair loader name which will be used to load
the raw text data.
-u USING, --using USING
A colon (:) separated column indexes. It is used for
limiting the reading columns.
-a, --average Calculate the average value of the specified data.
-s, --no-strict Do not strict the wavelength range into 190-240 .
-o OUTPUT, --output OUTPUT
A output filename. The default is "contin-cd.in".
Experimental properties:
-n NUMBER, --number NUMBER
The number of residues (amino acids) in the sample.
-m MOLECULAR_WEIGHT, --molecular-weight MOLECULAR_WEIGHT
A molecular weight of the sample in kDa (=kg/mol).
-c CONCENTRATION, --concentration CONCENTRATION
A concentration of the sample in g/L. See --molar-
concentration as an alternative.
--molar-concentration MOLAR_CONCENTRATION
A molar concentration of the sample in mol/L. It is
used as an alternative option of --concentration.
-L LENGTH, --length LENGTH
A light pathway length (cuvette length) in centimeter
Preference
You can create configure file as ~/.config/txt2contincd/txt2contincd.cfg (Linux), ~/.txt2contincd.cfg (Mac), or %APPDATA%\txt2contincd\txt2contincd.cfg (Windows).
The default preference is equal to the configure file as below:
[default] parser = 'parsers.PlainParser' loader = 'loaders.PlainLoader' using = None average = False strict = True output = 'contin-cd.in' [experiment] number = None molecular_weight = None concentration = None molar_concentration = None length = None
I don’t use Microsoft Windows so the location of the configure file in Windows might be wrong. Let me know if there are any mistakes.
Metadata
Release files for txt2contincd 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| txt2contincd-0.1.0.tar.gz | 16.3 kB | Without provenance | Details |
Release files / txt2contincd-0.1.0.tar.gz
| Download URL | txt2contincd-0.1.0.tar.gz |
|---|---|
| Size | 16.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
c0f489841ff38237f1dacecc8f60357bbc8a7b6e43bb5f78d8cff0b43f39d6ef
|
|
BLAKE2b-256 checksum How to use checksums |
c4c93ed6082beb0cb1bf2fc2966d8d5313cc6aa10aac543caa4a467bef8bc901
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
Security for txt2contincd 0.1.0
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.