This tool harnesses the interactive Python framework Dash to allow users analyzing their sequence-based data. Interactive plots and widgets are utilized to filter the data and discover sequence patterns. A web-based Dash application will be initialized to share the data with collegues over the web browser.
To use the app follow these instructions:
Cloning the source code to your local directory:
git clone https://github.com/lehecht/k-mer-Dash.git
Or downloading the source code from: https://github.com/lehecht/k-mer-Dash
k-mer Dash runs with python 3.6. It is recommended to use a virtual environment.
The app makes use of external python-libraries, which need to be installed. Setting up libraries, use:
pip3 install -r requirements.txt
Additionally ClustalW needs to be installed.
A system wide installation can be accomplished by (debian and ubuntu):
sudo apt install clustalw
Or by downloading from: http://www.clustal.org/clustal2/
| parameter | Description |
|---|---|
| -fs FILES | List of space separated FASTA formatted files with DNA nucleotide-sequences of equal length |
| -f FILE | Single FASTA formatted file with DNA nucleotide-sequences |
| -d DIR | Directory of FASTA formatted files with DNA nucleotide-sequences of equal length |
| -sfs FILES | List of space separated FASTA formatted file with element-strings of equal length |
| -sf FILE | Single FASTA formatted file with element-strings |
| -sd DIR | Directory of FASTA formatted files with element-strings of equal length |
| -k INT | Length of generated k-mer (substrings) |
| -p INT | Highlighted position in sequence, e.g. center of a binding site |
| -t INT | Number of highest k-mer occurrences used for visualization |
| -c STR | Boolean starting program with GUI or on command-line |
| -pt INT | Custom port (only for GUI) |
On valid input parameters a Dash application is started to explore various k-mer visualizations and statistics. It visualizes data as diagrams (Scatterplot, PCA, RNA Secondary Structure Heatmap) and tables (k-mer Frequency-table, Alignment-table).
In order to use the application via command-line only requires to set 'console' parameter -c on True.
Mandatory parameters are -f and -k. All other parameters are optional.
For commandline-mode only one single file can be evaluated.
The git project contains failsafe examples of generated fasta file in the example folder. An invokation to display the data on the console is listed below:
python3 main.py -f example/example1.fa -k 4 -t 5 -c True
Alignment of Top-kmere created with ClustalW
(for more information, see: http://www.clustal.org/clustal2/)
File: example1.fa
-TATA-
ATAT--
--AAAA
ACCC--
-CCCC-
AACC--
Options:
k: 4, peak: None, top: 5, files: ['example/example1.fa']
k-Mer Frequency File
AAAA 4 example1.fa
AACC 2 example1.fa
ACCC 2 example1.fa
CCCC 2 example1.fa
TATA 2 example1.fa
ATAT 2 example1.fa
Starting the application with the interactive user interface requires no parameter except for at least two FASTA-Files (-fs or -d). Port number (-pt) and structural data (-sd, -sfs, -sf) are optional.
python3 main.py -d example/ -sd example/exampleStrucData
PCA: Principle component analysis of selected files
Scatterplot: k-mer frequencies of selected files are visualized as point in coordinate system
RNA-Structure Heatmap: Arbitrary RNA-Structure based on alphabet of element-string. For example if alphabet contains no "i" or "I" (internal-loops), no internal-loops are visualized. For heatmap highlighting k-mer frequency is used. Only first occurrence of k-mer in RNA-Structure is highlighted. K-mer frequencies of element-sequences can be normalized by using 'extended options'-button.
Menu:
-
'Select-Files'-Dropdown: Select two files of given list of files, for which analysis should be done
-
Slider/Dropdown: Set values for their property
-
'Highlighted-Feature'-Dropdown: Change coloring in PCA plots from k-mer-frequency ('Frequency') to frequency of specific nucleotide (#T, #A, #C or #G) in k-mer
-
'Extended Options'-Button: Change options for RNA-Structure-Heatmap visualization
Tables:
-
Export-Button: Export tables to csv-file
-
Frequency-Table-Sorting: Sort by k-mers, frequency or filename
Plots:
- Dash-Features: Image-download, Zooming, Scaling, etc...
Tooltips: Information boxes for some GUI elements
MIT
ClustalW: Julie D. Thompson, Desmond G. Higgins, Toby J. Gibson, CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice, Nucleic Acids Research, Volume 22, Issue 22, 11 November 1994, Pages 4673–4680, https://doi.org/10.1093/nar/22.22.4673