Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Dec;49(12):1779-1784.
doi: 10.1038/ng.3984. Epub 2017 Oct 30.

Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells

Robin M Meyers  1 , Jordan G Bryan  1 , James M McFarland  1 , Barbara A Weir  1 , Ann E Sizemore  1 , Han Xu  1 , Neekesh V Dharia  1   2   3   4 , Phillip G Montgomery  1 , Glenn S Cowley  1 , Sasha Pantel  1 , Amy Goodale  1 , Yenarae Lee  1 , Levi D Ali  1 , Guozhi Jiang  1 , Rakela Lubonja  1 , William F Harrington  1 , Matthew Strickland  1 , Ting Wu  1 , Derek C Hawes  1 , Victor A Zhivich  1 , Meghan R Wyatt  1 , Zohra Kalani  1 , Jaime J Chang  1 , Michael Okamoto  1 , Kimberly Stegmaier  1   2   3   4 , Todd R Golub  1   2   3   4   5 , Jesse S Boehm  1 , Francisca Vazquez  1   2 , David E Root  1 , William C Hahn  1   2   4   6 , Aviad Tsherniak  1
Affiliations

Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells

Robin M Meyers et al. Nat Genet. 2017 Dec.

Abstract

The CRISPR-Cas9 system has revolutionized gene editing both at single genes and in multiplexed loss-of-function screens, thus enabling precise genome-scale identification of genes essential for proliferation and survival of cancer cells. However, previous studies have reported that a gene-independent antiproliferative effect of Cas9-mediated DNA cleavage confounds such measurement of genetic dependency, thereby leading to false-positive results in copy number-amplified regions. We developed CERES, a computational method to estimate gene-dependency levels from CRISPR-Cas9 essentiality screens while accounting for the copy number-specific effect. In our efforts to define a cancer dependency map, we performed genome-scale CRISPR-Cas9 essentiality screens across 342 cancer cell lines and applied CERES to this data set. We found that CERES decreased false-positive results and estimated sgRNA activity for both this data set and previously published screens performed with different sgRNA libraries. We further demonstrate the utility of this collection of screens, after CERES correction, for identifying cancer-type-specific vulnerabilities.

PubMed Disclaimer

Conflict of interest statement

Competing financial interests

W.C. Hahn reports receiving a commercial research grant from Novartis and is a consultant/advisory board member for the same as well as for KSQ Therapeutics. No potential conflicts of interest were disclosed by the other authors.

Figures

Figure 1
Figure 1. Genomic copy number confounds the interpretation of CRISPR-Cas9 loss-of-function proliferation screens of cancer cell lines
(a) Screen quality for each cell line in the panel (n=342), as measured by area under the receiver operating characteristic curve (AUC) in discriminating between predefined sets of common core essential and nonessential genes. (b) The depletion of sgRNAs is regressed against the number of perfect-match genomic cut sites using a simple saturating linear fit, which is plotted for each cell line, colored by lineage, and scaled such that the median of sgRNAs _targeting cell-essential genes is at −1, marked by a dashed line. (c) Genes are ranked by the mean depletion of _targeting sgRNAs (average guide score) and plotted for an example cell line. Values of 0 and −1 represent the median scores of nonessential and cell-essential genes, respectively, indicated by dashed lines. Below, depletion ranks of genes involved in fundamental cell processes and genes at various ranges of CN amplification are shown. (d) The median and interquartile range (IQR) of depletion ranks for the 100 most amplified genes per cell line are plotted. Color indicates mean amplification level of these genes. The gray-shaded area indicates the IQR of all genes screened.
Figure 2
Figure 2. Schematic of the CERES computational model
As input, CERES takes sgRNA depletion and CN data for all cell lines screened. During the inference procedure, CERES models the depletion values as a sum of gene-knockout and copy-number effects, multiplied by a guide activity score parameter. CERES then outputs the values of the parameters that produce the highest likelihood of the observed data under the model.
Figure 3
Figure 3. CERES corrects the copy-number effect and improves the specificity of fCRISPR-Cas9 essentiality screens while preserving true gene dependencies
(a) Boxplots of gene dependency scores are shown across CN for uncorrected average guide scores and CERES gene dependency scores. Data are scaled as in Fig. 1c. (b) The recall of cell-essential genes at a 5% FDR of nonessential genes is plotted for each cell line before (red) and after (blue) CERES correction. Precision-recall curves are inset for example cell lines with poor recall (bottom left) and good recall (top right) before CERES correction. (c) An example amplified region on chromosome 12p is shown for the DAN-G pancreatic cell line. The top track represents CN with amplifications shown in red. The middle track and bottom tracks show the average guide score and CERES score, respectively, for each gene in this region. The purple line is representing the median value in each CN segment. KRAS is highlighted in orange. (d) KRAS gene dependency and CN are shown for all cell lines after CERES correction, with mutant KRAS lines in orange.
Figure 4
Figure 4. CERES estimates guide activity scores for each sgRNA
(a) sgRNAs are binned into groups with high (0.9–1), moderate (0.2–0.9), and low (0–0.2) guide activity scores. The compositions of guide activity scores are shown for the set of screens performed with the GeCKOv2 and the Avana sgRNA libraries. (b) For the set of 4,770 sgRNAs shared between the GeCKOv2 and Avana libraries, sgRNAs are ranked by guide activity scores in each dataset and are plotted against each other, with darker blue representing a greater density of sgRNAs. (c) sgRNAs are binned by predicted on-_target activity using the Doench-Root score, and the composition of CERES-estimated guide activity scores is shown for each dataset.
Figure 5
Figure 5. CERES reduces false positive differential dependencies
(a) The percentage of genes on amplified regions (CN > 4) below a given differential dependency threshold is plotted for the uncorrected average guide score in red and the CERES gene dependency score in blue. (b) The percentage of unexpressed genes (log2RPKM < −1) below a given differential dependency score is plotted as in (a).
Figure 6
Figure 6. CERES reduces false positives among lineage-specific differential dependencies due to recurrently amplified chromosome arms
(a) The distributions of differential dependencies in breast lines are plotted red for genes on chromosome 8q (commonly gained in breast tumors) and black for all other genes. Below, the differential dependency of each gene is plotted against the FDR-corrected p-value, calculated from a student’s t-test, with colors as above. The dashed line represents an FDR of 5%. (b) Data are shown for CERES-inferred gene effects as in (a). (c) The percentages of lineage-specific differential dependencies (FDR < 0.05) that are on recurrently amplified chromosome arms are shown, before and after CERES correction.
See this image and copyright information in PMC

Comment in

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Wang T, et al. Identification and characterization of essential genes in the human genome. Science. 2015;350:1096–1101. - PMC - PubMed
    1. Hart T, et al. High-Resolution CRISPR Screens Reveal Fitness Genes and Genotype-Specific Cancer Liabilities. Cell. 2015;163:1515–1526. - PubMed
    1. Aguirre AJ, et al. Genomic Copy Number Dictates a Gene-Independent Cell Response to CRISPR/Cas9 _targeting. Cancer Discov. 2016;6:914–929. - PMC - PubMed
    1. Munoz DM, et al. CRISPR Screens Provide a Comprehensive Assessment of Cancer Vulnerabilities but Generate False-Positive Hits for Highly Amplified Genomic Regions. Cancer Discov. 2016;6:900–913. - PubMed
    1. Cheung HW, et al. Systematic investigation of genetic vulnerabilities across cancer cell lines reveals lineage-specific dependencies in ovarian cancer. Proc Natl Acad Sci USA. 2011;108:12372–12377. - PMC - PubMed
  NODES
twitter 2