Decipher Immunosignature Code 07-09-2014d0907

2015-01-13

Decipher Immunosignature Code 07-09-2014d0907

cim_logistics of getting scores for fragments in a database for a set of epitopes 08-01-2014d0950

future code changes

I want to be able to list what the epitope is
I want to be able to view detailed output of alignments (optional parameter)
I want to be able to remove redundant fragments from a database and results
fix blast error that occurs sometimes when searching for epitopes from the random peptides
make it so that only epitopes within a certain size are returned and remove duplicates from the final population
setup the program so that it can go straight from finding the epitopes to searching a database
clean up code so that unused code and classes are removed
adjust code so that when epitopes are found, the population size, number of evolution cycles, and glam2 threshold are automatically output as well

Stages for running program.

Perform genetic algorithm with 100 peptides from t-test from immunosignature
Prepare the database you want to analyze by creating the curated sequences.
Break the found motifs up into all 3, 4, and 5 mer epitopes and perform a genetic algorithm to match these fragments with 30 aa fragments in a database. Do this across many computers.
Combine the results from all of the different 30aa fragments analyzed.

Custom Epitope Group With Protein Fitness Evaluator code to evaluate in a parallel manner
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-06-2014d1041\CustomEpitopeGroupWithProteinFitnessEvaluator
-Is this code faster? Not really.
--cim_time evaluation 08-06-2014d2155

Instance of code on other computers

C:\ativ6_storage\2014\8-24-14\decipher test 8-10-14
/home/owner/ratel_ultra_storage/2014/8-24-14/Decipher/DecipherImmunosignature

Versions of Decipher Immunosignature Code

current code 09-11-2014d2007
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\09-11-2014d2008\Decipher Immunosignature\DecipherImmunosignature
fixed a bug in the code so that protein sequences only containing letters from DNA can be analyzed by adding the "--seqtype=protein" in the clustal command line part of the code
-see C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\09-09-2014d0915\clustalo test
current code 09-04-2014d1058
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\09-04-2014d1059\09-04-2014d1059\Decipher Immunosignature
current code 09-01-2014d1256 .. this version of the code did a fairly good job of screening a database .. version before adding code so that only the top four mers are used rather than all of the four mers that pass a threshold
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\09-01-2014d1257\09-01-2014d1257\Decipher Immunosignature\DecipherImmunosignature
-some results
--C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\09-02-2014d1650
current code 08-27-2014d1122. . right when I first fixed the following error: collections sort throws comparison method violates its general contract exception
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-27-2014d1011\08-27-2014d1123\Decipher Immunosignature\DecipherImmunosignature
current code 08-27-2014d1012
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-27-2014d1011\Decipher Immunosignature\DecipherImmunosignature
current code 08-26-2014d0930. . cod before modifying alignment evaluation so that broken epitopes were not considered
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-26-2014d0932\Decipher Immunosignature\DecipherImmunosignature
current code 08-25-2014d0955. . code before changing epitope database search so that there is a max number of threads
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-25-2014d0953\Decipher Immunosignature\DecipherImmunosignature
current code 08-20-2014d1647. . version before changing database search so that all four mair pairs are searched first for each protein fragment
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-20-2014d0937\08-20-2014d1647\DecipherImmunosignature"
current code 08-20-2014d0936. . version before modifying genetic algorithm so that previously encountered chromosomes are not re-evaluated.
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-20-2014d0937\DecipherImmunosignature"
current code 08-19-2014d0939. . version before modifying program so it could work a little better in Linux.
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-19-2014d0923\08-19-2014d0941\DecipherImmunosignature
current code 8-19-14d0922. The parallel multi-thread method for finding epitope groups seems to be working in this version of the code.
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-19-2014d0923\DecipherImmunosignature"
current code 08-17-2014d1059.
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-17-2014d1100\DecipherImmunosignature
current code 08-09-2014d1125. Code before changing evolve epitope group code so that it could handle multiple threads. 08-17-2014d1059

C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-09-2014d1125\DecipherImmunosignature

current code pre 08-09-2014d1124. . this is version of code before I started to change things so that an epitope group with a protein is not evaluated so many times, and I also wanted to try the genetic algorithm with multiple threads.
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-05-2014d1134\new\DecipherImmunosignature"
This version of the code has a EpitopeGroupWithProteinMatchesFromOmega class which assigns a score based on the percentage of the max possible score. I wanted to save this before I modified the class to score in my new way which emphasizes the importance of lots of consecutive matches.
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\08-05-2014d1134\DecipherImmunosignature"
I will start using multiple instances of eclipse with the same code at a location on the hard drive (not in eclipse). .version of code which uses blast to make sure current motif is not similar to one of the previously collected motifs. I also just started adding some age code to the chromosomes so that chromosomes that are too old will be discarded.
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\07-18-2014d0956\Decipher Code versions\07-18-2014d0957\DecipherImmunosignature"
--length of path is 148 characters so this should be acceptable
working version which checks if a chromosome has a score greater than 33 from glam2. If it does, then if it matches an existing motif already found, then give this chromosome a low fitness score. This program found the HEE motif and SS motif in the top 10 best motifs. I would like to adjust the program though so that a motif is checked against the existing best collected motifs. If it is very similar to an existing motif, then this motif should get a low fitness score so that new motifs can be found. Maybe I will use blast to perform this comparison.
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\07-11-2014d0935\DecipherImmunosignature
--some results obtained with this code
---C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\07-08-2014d0933\decipher work 07-08-2014d0933\evolution with avoidance of high scoring already found motifs 07-11-2014d0950
working version which can take a list of peptides and evolve them starting with a population with no more than an allowed number of positive genes (the min is specified as well (to the value of 2)) 07-09-2014d1050
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\07-09-2014d0909\07-09-2014d1050\DecipherImmunosignature"
working version 07-09-2014d0908
-C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\07-09-2014d0909\DecipherImmunosignature