marcottelab/FluorosequencingSimulation

Name: FluorosequencingSimulation

Owner: The Marcotte Lab

Description: null

Created: 2014-12-16 22:14:54.0

Updated: 2014-12-17 16:37:47.0

Pushed: 2014-12-17 16:37:47.0

Homepage: null

Size: 174

Language: C

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

FluorosequencingSimulation

INTRODUCTION

---

The scripts (a) seqmodlibMP.py and (b) randsiggen.c codes for the steps in the Monte Carlo simulation detailed in the paper. The two scripts form the underlying code and a wrapper code must be applied according to the parallelized computing infrastructure used.

System Requirements

---

1. Python 2.7
2. Standard C compiler
3. High Performance Computing (HPC) infrastructure capable of parallel computing.
4. libGSL
5. SMFT Merseinne Twister

INSTALLATION

---

Unzipr or Untar the files into a folder (preferably added to the PYTHONPATH) *Ipython IDE may be installed to write and execute the wrapper codes randsiggen.c needs to know where libGSL and SMFT header files are

PROGRAMME EXECUTION

---

(a) PREPROCESSING PROTEOME FILE

1. Proteome of choice can be downloaded from a number of publicly available databases like Uniprot (http://www.uniprot.org/) etc user defined. 
2. The file is parsed to yield a python dictionary in the form : 
     {
           'PROTEIN NAME 1': 'AMINO ACID SEQUENCE 1',
           'PROTEIN NAME 2': 'AMINO ACID SEQUENCE 2',
         ...}
3. The file is stored as a pkl file. 

(b) WRAPPER CODE

For example: 

    import seqmodlibMP
    proteome = seqmodlibMP.load_proteome(proteome_pkl_File) #proteome pkl file is loaded
    cleaved = seqmodlibMP.cleave(proteome,'E') # This cleaves all peptides after 'E' (representing the GluC cleavage)
    attached = seqmodlibMP.attach(cleaved,'C') # This retains only 'C' containing peptides 

    trie = seqmodlibMP.monte_carlo_trie(attached, 
                    p=0.9,    #90% Edman efficiency
                    b=0.0033, #photobleaching rate constant of 0.0033
                    u=0.20,   #fluorophore failure rate of 20%
                    windows={'K':range(1,30)}, #Considers upto 30 experimental cycles with the label on 'K'
                    sample_size = 1000, #Represents the simulation depth
                    random_seed=random.random(), #Generates the appropriate random model for the error sources
                    silent=True #True turns off the print statement. Useful for debugging
                        )

    trie.find_uniques()     # Trie Analysis is done by calling 

NOTE *This module also needs to be implemented in a HPC (see system requirements). The function needs a wrapper that partitions the resultant trie over a number of multiterabyte-sized tries. *

(c) RESULT FILE

The end result file has a mapping of a fluorosequence to the tally of source peptides generating it
For example: 
    xxKxxk : (Protein A, 20),(Protein B, 8), ...
    xKxxK  : (Protein C, 200), (Protein A, 20), ...
    ...

(d) VISUALIZING DATA

All the data were graphed for visualization using standard python matplotlib library functions (http://matplotlib.org/)

AUTHOR

---

Alexander Boulgakov (alexander[DOT]boulgakov{AT}utexas[DOT]edu)