| MAGeCK [54–56] |
Negative binomial model — based analysis of genome-wide CRISPR–Cas9 KO screens for prioritizing sgRNAs, genes and pathways. |
Python, R |
| HiTSelect [57] |
Uses Poisson distribution to evaluate sgRNAs and stochastic multiobjective ranking method to generate gene-level statistics. |
Matlab |
| ScreenBEAM [58] |
Bayesian hierarchical (multilevel) model to directly assess gene-level activity from all relevant measurements. |
R |
| STARS [12] |
Gene-ranking algorithm for genetic perturbation screens — gene scores are computed using the probability mass function of a binomial distribution. |
Python |
| BAGEL [59,60] |
Bayesian analysis for identifying essential genes from pooled screens, based on core essential and nonessential gene sets. |
Python |
| CaRpools [61] |
A pipeline for end-to-end analysis of pooled CRISPR/Cas9 screening data. Including in-depth analysis of screening quality and sgRNA phenotypes. |
R |
| CasTLE [62] |
Maximum likelihood estimator and empirical Bayesian framework to account for multiple sources of variability, including reagent efficacy and off-target effects for the analysis of large-scale genomic perturbation screens. |
Python |
| CERES [5] |
A method to estimate gene dependency from essentiality screens while computationally correcting the copy number effect, therefore enabling unbiased interpretation of gene dependency at all levels of copy number. |
R |
| ENCoRE [63] |
Workflow for NGS to CRISPR gene results. |
Java |
| PBNPA [64] |
Permutation-based non-parametric analysis, which computes P-values at the gene level by permuting sgRNA labels, therefore avoids restrictive distributional assumptions. |
R |
| CRISPhieRmix [65] |
Broad-tailed null distribution is fit using negative control sgRNAs. Then, a mixture distribution is fit on all sgRNAs, ignoring gene identities. Lastly, using the mixture distribution the false discovery rate for each gene is calculated. |
R |
| CB2 [66] |
Beta-binomial model with a modified Student's t-test to measure differences in sgRNA levels, followed by Fisher's combined probability test to estimate the gene-level significance. |
R |
| JACKS [67] |
Bayesian method that jointly analyzes screens performed with the same library and assigns a gene P-value based on empirically derived null distribution based on essentiality scores in a known set of negative control genes. |
Python |
| DrugZ [68] |
Identifies synergistic and suppressor drug-gene interactions from CRISPR-based chemogenetic screens. |
Python |
| Gscreend [69] |
Mixture of a parametric null distribution is used to calculate P-value for every sgRNA, and robust rank aggregation (RRA) algorithm is used to aggregate and score the data on gene-level. |
R |
| CRISPRcleanR [70] |
Unsupervised copy number correction of gene-independent responses in genome wide CRISPR KO screens based on circular binary segmentation algorithm. |
Python, R |
| CRISPy [71] |
Supervised copy number correction of gene-independent effects, which uses Gaussian processes regression to model non-linear effects between the segment copy number ratio and CRISPR fold changes. |
R |
