Python Codon Adaptation Index¶
An implementation of Sharp and Li’s 1987 formulation of the codon adaption index.
Installation¶
This module is available from PyPI and can be downloaded with the following command:
$ pip install CAI
To install the latest development version:
$ pip install git+https://github.com/Benjamin-Lee/CodonAdaptationIndex.git
Quickstart¶
Finding the CAI of a sequence is easy:
>>> from CAI import CAI
>>> CAI("ATG...", reference=["ATGTTT...", "ATGCGC...",...])
0.24948128951724224
Similarly, from the command line:
$ CAI -s sequence.fasta -r reference_sequences.fasta
0.24948128951724224
Determining which sequences to use as the reference set is left to the user, though the HEG-DB is a great resource of highly expressed genes.
Contributing and Getting Support¶
If you encounter any issues using CAI, feel free to create an issue.
To contribute to the project, please create a pull request. For more information on how to do so, please look at GitHub’s documentation on pull requests.
Citation¶
Lee, B. D. (2018). Python Implementation of Codon Adaptation Index. Journal of Open Source Software, 3 (30), 905. https://doi.org/10.21105/joss.00905
@article{Lee2018,
doi = {10.21105/joss.00905},
url = {https://doi.org/10.21105/joss.00905},
year = {2018},
month = {oct},
publisher = {The Open Journal},
volume = {3},
number = {30},
pages = {905},
author = {Benjamin D. Lee},
title = {Python Implementation of Codon Adaptation Index},
journal = {Journal of Open Source Software}
Contact¶
I’m available for contact at benjamin_lee@college.harvard.edu.
Reference¶
Sharp, P. M., & Li, W. H. (1987). The codon adaptation index–a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Research, 15(3), 1281–1295.
Table of Contents¶
Usage¶
Basic Usage¶
As covered in Quickstart, the basic CAI() function is fast and
easy. Simply import it and get to your science. Note that it also plays nicely
with Biopython Seq objects:
>>> from CAI import CAI
>>> from Bio.Seq import Seq
>>> CAI(Seq("AAT"), reference=[Seq("AAC")])
0.5
The CLI is equally easy to use. For example, to find the CAI of the native GFP gene with respect to the highly expressed genes in E. coli, only one command is required:
$ CAI -r example_seqs/ecol.heg.fasta -s example_seqs/gfp.fasta
0.3753543123685772
Note
Both CAI and cai are valid commands.
More example sequences can be found in the example_seqs directory on GitHub.
Advanced Usage¶
If you have already computed the weights or RSCU values of the reference set,
you can supply CAI() with one or the other as arguments. They must be
formatted as a dictionary and contain values for every codon.
To calculate RSCU without calculating CAI, you can use RSCU(). RSCU()’s only
required argument is a list of sequences.
Similarly, to calculate the weights of reference sequences, you can use
relative_adaptiveness(). relative_adaptiveness() takes either a list of
sequences as the sequences parameter or a dictionary of RSCUs as the RSCUs
parameter.
Warning
If you are computing large numbers of CAIs with the same reference
sequences, first calculate their weights with relative_adaptiveness()
and then pass that to CAI() to eliminate redundant computation.
So, to modify the example in Quickstart:
>>> from CAI import CAI, relative_adaptiveness
>>> sequences=["ATGTTT...", "ATGCGC...",...]
>>> weights = relative_adaptiveness(sequences=sequences)
>>> CAI("ATG...", weights=weights)
0.24948128951724224
These are exactly equivalent:
>>> assert CAI("ATG...", weights=weights) == CAI("ATG...", reference=sequences)
True
except the former will be faster if you’re using the same weights repeatedly.
Other Genetic Codes¶
All functions in CAI support an optional genetic_code parameter, which is set
by default to 11 (the standard genetic code).
In the CLI, there is an optional “-g” parameter that changes the genetic code:
$ CAI -s sequence.fasta -r reference_sequences.fasta -g 22
0.25135779681923687
API Reference¶
-
RSCU(sequences, genetic_code=11)¶ Calculates the relative synonymous codon usage (RSCU) for a set of sequences.
RSCU is ‘the observed frequency of [a] codon divided by the frequency expected under the assumption of equal usage of the synonymous codons for an amino acid’ (page 1283).
In math terms, it is
\[\frac{X_{ij}}{\frac{1}{n_i}\sum_{j=1}^{n_i}x_{ij}}\]“where \(X\) is the number of occurrences of the \(j\) th codon for the \(i\) th amino acid, and \(n\) is the number (from one to six) of alternative codons for the \(i\) th amino acid” (page 1283).
Parameters: Returns: The relative synonymous codon usage.
Return type: Raises: ValueError– When an invalid sequence is provided or a list is not provided.
-
relative_adaptiveness(sequences=None, RSCUs=None, genetic_code=11)¶ Calculates the relative adaptiveness/weight of codons.
The relative adaptiveness is “the frequency of use of that codon compared to the frequency of the optimal codon for that amino acid” (page 1283).
In math terms, \(w_{ij}\), the weight for the \(j\) th codon for the \(i\) th amino acid is
\[w_{ij} = \frac{\text{RSCU}_{ij}}{\text{RSCU}_{imax}}\]where “\(\text{RSCU}_{imax}\) [is] the RSCU… for the frequently used codon for the \(i\) th amino acid” (page 1283).
Parameters: Note
Either
sequencesorRSCUsis required.Returns: A mapping between each codon and its weight/relative adaptiveness.
Return type: Raises: ValueError– When neithersequencesnorRSCUsis provided.ValueError– SeeRSCU()for details.
-
CAI(sequence, weights=None, RSCUs=None, reference=None, genetic_code=11)¶ Calculates the codon adaptation index (CAI) of a DNA sequence.
CAI is “the geometric mean of the RSCU values… corresponding to each of the codons used in that gene, divided by the maximum possible CAI for a gene of the same amino acid composition” (page 1285).
In math terms, it is
\[\left(\prod_{k=1}^Lw_k\right)^{\frac{1}{L}}\]where \(w_k\) is the relative adaptiveness of the \(k\) th codon in the gene (page 1286).
Parameters: Note
One of
weights,referenceorRSCUsis required.Returns: The CAI of the sequence.
Return type: Raises: TypeError– When anything other than one of either reference sequences, or RSCU dictionary, or weights is provided.ValueError– SeeRSCU()for details.KeyError– When there is a missing weight for a codon.
Warning
Will return nan if the sequence only has codons without synonyms.
CLI Reference¶
$ CAI --help
Usage: CAI [OPTIONS]
Options:
-s, --sequence FILE The sequence to calculate the CAI for.
[required]
-r, --reference FILE The reference sequences to calculate CAI
against. [required]
-g, --genetic-code INTEGER The genetic code to use. Defaults to 11.
--help Show this message and exit.
License¶
This software is licensed under the MIT License. If you’re unfamiliar with software licenses, here is a handy summary of the license.
For reference, the license is reproduced below:
MIT License
Copyright (c) 2017 Benjamin Lee
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.