Manual

User Manual:

Open the PDF directly: View PDF PDF.
Page Count: 2

DownloadManual
Open PDF In BrowserView PDF
Framework for functional profile prediction
User Guide
Requirements:
To run this tool you have to have Python 3 installed. If you want to use the bash script
intended for trying out the tool, you also have to have bash command line.
Input:
A table with OTU identifiers and abundances in the given sample, closed-reference picked
against Greengenes, format .tab. ​(Provided in the data/test directory as sample*.tab)
Output:
KO profile table. ​(Expected output to compare accuracy provided in the data/test directory
as expected*.tab)
Running the tool:
The easiest way to run the tool is from the command line from folder framework:
$ ./test.sh python-alias [-m method-name] [-i input-filename] [-o output-filename]
Running the script with ​-h parameter will show help, including description of parameters and
available methods for functional prediction.
Available settings:
If you want to change other parameters than input, output and method, you need to change
them in file settings.py. ​(For the purpose of trying out this tool, you do not have to change
anything, default settings will be used)
Following is the description of all parameters that can be changed in settings.py.
Global settings (same for each method):
● KO_PROFILE_FILENAME​ - file with reference ko profiles
● INPUT_FILENAME​ - input sample
● UNKNOWN_METHOD​ - method to use for prediction
● RESULT_FILENAME​ - result file
● COUNT_16S_FILENAME​ - file with 16s counts
Methods for prediction and specific settings:
● alignment_simple
○ SIMILAR​ - how many similar sequences should be averaged
○ SCORE_FILENAME​ - distance matrix file
● random

○ SIMILAR​ - how many similar sequences should be averaged
● tree
○ TREE_FILE ​- file with phylogenetic tree
● treshold
○ TRESHOLD​ - similarity treshold in percent
● weighted
○ MAX_DIFF​ - maximal difference to consider sequences as similar
○ SCORE_FILENAME​ - distance matrix file
○ SIMILAR​ - how many similar sequences should be averaged
● regression_by_sequence
○ RESTS​ - how much of the reference table is known
Source Exif Data: 
File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.5
Linearized                      : Yes
Producer                        : Skia/PDF m76
Page Count                      : 2
EXIF Metadata provided by EXIF.tools

Navigation menu