import numpy as NP

import pylab as PL

import FlowCytometryTools

from FlowCytometryTools import FCMeasurement

import glob as glob

# DECOMPOSER LE CODE EN PLUSIEURS PETITES PARTIES ET ANALYSER CHACUNE POUR TROUVER L'ERREUR

# programme : utiliser le flow Cytometer pour les nouveaux genes

# grow PRY3 in a tube in the AFTERNOON

# Take new plate ET FAIRE LE GENE QUI NA PAS GRNDI : 10 AM

# DO fits : fit a line to the data

# do the linear regression beteween the GFP and the size scatter

# regrow it , repull it another time MONDAY , le gene qui ne marchait pas

# Il m'a envoyé des files a regarder

# FUNCTION: Loads data and converts to numpy array

# datafile=r'Data.001'

def load_file(File_name):

return current_data,sample_id

# This function takes our current data file and erases some values that are above a threshold we fixed in

# order to get rid of all the supplementary signals the flow cytometer takes into account

# and that we do not need

def gate_cells(cells,fsc_thr,ssc_thr,cut_max = True, cut_GFP_max = True):

return cells5

# Writing a function that will directly return all the sample IDs, the raw data and the filtered data for our files (so we won't be doing it with each file)

def filter_all_data(fsc_thr, ssc_thr, max_treat = True):

return all_data, strain_list # We return a dictionary that contains all the data for every file and a list that contains all the gene names that are present in the files.

# FUNCTION: Bin data by size

# Takes all data and bins across range of values. Constant size bins enforced to permit

# background subtraction bin-wise.

# It bins the data meaning for x it takes the middle and for y it takes the mean , so for a range of points we only select one point , and it gives us a better understanding of the data , and a clearer vieuw of the curve.

def plot_size_bins(x, y, n_bin=100,plot_range = [0, 1023]): # What is x? What is y?

return NP.array(binned_data)

# This function bins the data for the SSC and the GFP and puts the binned data in a new dictionary , so now we have the filtered data that we want

def bin_all(data):

return binned_data

# These functions's goal is to substract the unnecessary things from our plot to be able to plot the filtered data

# For us ,the bg_strain is 'BY'

# Definition of the BG_strain?

# The strain list contains the names of the genes that we have in our files

def bg_sub_wrap(raw_data, binned_data, bg_strain, strain_list):

return new_binned_data, new_raw_data

#?

def bg_sub(raw_data, binned_data, bg_strain, sample, strain_list):

return all_vals, all_vals_raw

# Plot background subtracted graphs.

# Takes all strains in current data from binned data and plots them individually. Saves

# plots as PDF.

def plot_save_data(proc_data, binned_data, file_type = 'pdf'): for sample in proc_data.keys()

temp_data = proc_data[sample]

temp_bin = binned_data[sample]

PL.plot(proc_data[sample][:,1], proc_data[sample][:,3], '.', markersize = 12, \

alpha = .05, label = "{0}".format(sample))

PL.plot(temp_bin[:,0],temp_bin[:,3],'-k', linewidth = 2)

PL.legend(loc = "upper left")

PL.xlabel("SSC (Volume)")

PL.ylabel("GFP Intensity (AU)")

PL.ylim(0,1050)

PL.xlim(0,1050)

savepath = '{0}.{1}'.format(sample,file_type)

PL.savefig(savepath)

PL.close()

# New raw data is processed data

def analyze_data(fsc_thr, ssc_thr, bg_strain):

print "Process complete."

# Code to do the analysis for the Flow Cytometry

# Monday : The FlowCytometer

# HMO1 and PRY3 are 2 genes we are going to test on the

# microscope

# MAKING CULTURES OF THEM

# Do culture of PRY 3 Monday night and watch them on the microscope

# The tube just like for flow Cytometry

# HMO1 is it cell cycle regulated? because we obtained a weird

# Cycle base

# PRY3 seems to be non SCALING !!!! We can see that the GFP and the side scatter : consstante , the amount does not depend on the volume of the cell

# We have 3 non scaling genes here for today which is great! Now we should just get some info on them ,

# then go test them on the microscope:

# Tomorrow let's find 3 others that could be non scaling, then do it all over again and see what I get.

# Et Tomorrow, microscope

# ADD PEX13 ON MY NEXT SET OF GENES

# ACE2 , APL3, MET30,PEX13 look NON scaling , have to do the fittings to make sure

# After bg substraction, before PLOTTING

# FIT

# HHO1 nn scaling gene we added from today's list, design PRIMAR

# VRP1 plot the linear regression

import sklearn

linear_regression(raw_data)

from sklearn.linear_model import LinearRegression

def linear_model(proc_data)

new_linear_data = {}

for sample in proc_data.keys()

lm=LinearRegression()

a=lm.fit(proc_data[sample][:,1], proc_data[sample][:,3])

b=lm.intercept

new_linear_data[sample]=a

new_linear_data[coeff]=b

return new_linear_data

# Take BACKGROUND SUBSTRACTED DATA ,

# Fit a linear function to data

# Linear regression

# perform linear regressio non SSC versus GFP

# store the linear regression in my dictionary raw data

# take the Y intercepts of that fit and divide by the mean GFP value and store that number

# FOR EACH GENE

# STORE THE FIT , the Y intercept.

# store something in a dictionary.

# Create a dictionary