diff --git a/Scripts/PrintStatus.py b/Scripts/PrintStatus.py index d2f755d..10d0c59 100755 --- a/Scripts/PrintStatus.py +++ b/Scripts/PrintStatus.py @@ -12,8 +12,8 @@ def PrintStatus_Header(): print('**************************************************') - print('Launching PinAPL-Py v2.911') - print('P. Spahn et al., UC San Diego (10/2020)') + print('Launching PinAPL-Py v2.9.2') + print('P. Spahn et al., UC San Diego (12/2020)') print('**************************************************') def PrintStatus_SubHeader(msg): diff --git a/Scripts/RankGenes_SigmaFC.py b/Scripts/RankGenes_SigmaFC.py index 40b6c02..8076438 100644 --- a/Scripts/RankGenes_SigmaFC.py +++ b/Scripts/RankGenes_SigmaFC.py @@ -29,11 +29,12 @@ def SigmaFC_Permutation(P): N0 = list(sg_table['control mean']); pi = 0 for j in range(k): - y = np.log10((Nx[j]+delta)/(N0[j]+delta)) # log fold-change - if y < FCmin: - u = np.exp(Lambda*(y-FCmin)) - else: - u = 1 + # compute fold-change weight (discontinued) +# y = np.log10((Nx[j]+delta)/(N0[j]+delta)) # log fold-change +# if y < FCmin: +# u = np.exp(Lambda*(y-FCmin)) +# else: + u = 1 pi_j = u * np.log10((Nx[j]+delta)/(N0[j]+delta)) pi = pi + pi_j #Pi = w[k]*pi # only w = id supported @@ -59,9 +60,9 @@ def compute_SigmaFC(sgRNAList): alpha_g = config['alpha_g'] ScreenType = config['ScreenType'] num_cores = multiprocessing.cpu_count() - global Lambda; Lambda = -np.log(0.1) # deprecation rate - global FCmin; # deprecation cutoff (log10 fold change). Set to 0 for no deprecation - FCmin = config['FCmin_SigmaFC'] +# global Lambda; Lambda = -np.log(0.1) # deprecation rate +# global FCmin; # deprecation cutoff (log10 fold change). DISCONTINUED +# FCmin = config['FCmin_SigmaFC'] global w; # reward function for numbers of signif. sgRNAs w = config['w_SigmaFC'] @@ -110,13 +111,13 @@ def compute_SigmaFC(sgRNAList): Nx = list(sg_table['counts']) N0 = list(sg_table['control mean']) pi = 0 - for j in range(k): - # compute fold-change weight - y = np.log10((Nx[j]+delta)/(N0[j]+delta)) # log fold-change - if y < FCmin: - u = np.exp(Lambda*(y-FCmin)) - else: - u = 1 + for j in range(k): + # compute fold-change weight (discontinued) +# y = np.log10((Nx[j]+delta)/(N0[j]+delta)) # log fold-change +# if y < FCmin: +# u = np.exp(Lambda*(y-FCmin)) +# else: + u = 1 pi_j = u * np.log10((Nx[j]+delta)/(N0[j]+delta)) pi = pi + pi_j else: