b788a707887a79434941ee37b1c5f5c496995c70 kuhn Thu Jan 20 07:14:25 2022 -0800 added definition of CFD score diff --git src/hg/makeDb/trackDb/crisprAll.html src/hg/makeDb/trackDb/crisprAll.html index 154f5e0..7db918b 100644 --- src/hg/makeDb/trackDb/crisprAll.html +++ src/hg/makeDb/trackDb/crisprAll.html @@ -1,210 +1,211 @@ <h2>Description</h2> <p> This track shows the DNA sequences targetable by CRISPR RNA guides using the Cas9 enzyme from <em>S. pyogenes</em> (PAM: NGG) over the entire $organism ($db) genome. CRISPR target sites were annotated with predicted specificity (off-target effects) and predicted efficiency (on-target cleavage) by various algorithms through the tool <a href="http://crispor.tefor.net/" target="_blank">CRISPOR</a>. Sp-Cas9 usually cuts double-stranded DNA three or four base pairs 5' of the PAM site. </p> <h2>Display Conventions and Configuration</h2> <p> The track "CRISPR Targets" shows all potential -NGG target sites across the genome. The target sequence of the guide is shown with a thick (exon) bar. The PAM motif match (NGG) is shown with a thinner bar. Guides are colored to reflect both predicted specificity and efficiency. Specificity reflects the "uniqueness" of a 20mer sequence in the genome; the less unique a sequence is, the more likely it is to cleave other locations of the genome (off-target effects). Efficiency is the frequency of cleavage at the target site (on-target efficiency).</p> <p>Shades of gray stand for sites that are hard to target specifically, as the 20mer is not very unique in the genome:</p> <table class="stdTbl" style="width:100%"> <tr><td style="width:50px; background-color:#969696"></td><td>impossible to target: target site has at least one identical copy in the genome and was not scored</td></tr> <tr><td style="width:50px; background-color:#787878"></td><td>hard to target: many similar sequences in the genome that alignment stopped, repeat?</td></tr> <tr><td style="width:50px; background-color:#505050"></td><td>hard to target: target site was aligned but results in a low specificity score <= 50 (see below)</td></tr> </table> <p>Colors highlight targets that are specific in the genome (MIT specificity > 50) but have different predicted efficiencies:</p> <table class="stdTbl" style="width:100%"> <tr><td style="width:50px; background-color:#000064"></td><td>unable to calculate Doench/Fusi 2016 efficiency score</td></tr> <tr><td style="width:50px; background-color:#FF7070"></td><td>low predicted cleavage: Doench/Fusi 2016 Efficiency percentile <= 30</td></tr> <tr><td style="width:50px; background-color:#FFFF00"></td><td>medium predicted cleavage: Doench/Fusi 2016 Efficiency percentile > 30 and < 55</td></tr> <tr><td style="width:50px; background-color:#00b300"></td><td>high predicted cleavage: Doench/Fusi 2016 Efficiency > 55</td></tr> </table><BR> <p> Mouse-over a target site to show predicted specificity and efficiency scores:<br> <ol> <li>The MIT Specificity score summarizes all off-targets into a single number from 0-100. The higher the number, the fewer off-target effects are expected. We recommend guides with an MIT specificity > 50.</li> <li>The efficiency score tries to predict if a guide leads to rather strong or weak cleavage. According to <a href="#References">(Haeussler et al. 2016)</a>, the <a href="https://portals.broadinstitute.org/gpp/public/analysis-tools/sgrna-design">Doench 2016 Efficiency score</a> should be used to select the guide with the highest cleavage efficiency when expressing guides from RNA PolIII Promoters such as U6. Scores are given as percentiles, e.g. "70%" means that 70% of mammalian guides have a score equal or lower than this guide. The raw score number is also shown in parentheses after the percentile.</li> <li>The <a href="https://www.crisprscan.org/">Moreno-Mateos 2015 Efficiency score</a> should be used instead of the Doench 2016 score when transcribing the guide in vitro with a T7 promoter, e.g. for injections in mouse, zebrafish or Xenopus embryos. The Moreno-Mateos score is given in percentiles and the raw value in parentheses, see the note above.</li> </ol> </p> <p>Click onto features to show all scores and predicted off-targets with up to four mismatches. The Out-of-Frame score by <a href="#References">Bae et al. 2014</a> is correlated with the probability that mutations induced by the guide RNA will disrupt the open reading frame. The authors recommend out-of-frame scores > 66 to create knock-outs with a single guide efficiently.<p> -<p>Off-target sites are sorted by the CFD score (<a href="https://www.nature.com/articles/nbt.3437" +<p>Off-target sites are sorted by the CFD (Cutting Frequency Determination) +score (<a href="https://www.nature.com/articles/nbt.3437" target="_blank">Doench et al. 2016</a>). The higher the CFD score, the more likely there is off-target cleavage at that site. Off-targets with a CFD score < 0.023 are not shown on this page, but are available when following the link to the external CRISPOR tool. When compared against experimentally validated off-targets by <a href="#References">Haeussler et al. 2016</a>, the large majority of predicted off-targets with CFD scores < 0.023 were false-positives. For storage and performance reasons, on the level of individual off-targets, only CFD scores are available.</p> <h2>Methods</h2> <h3>Relationship between predictions and experimental data</h3> <p> Like most algorithms, the MIT specificity score is not always a perfect predictor of off-target effects. Despite low scores, many tested guides caused few and/or weak off-target cleavage when tested with whole-genome assays (Figure 2 from <a href="#References">Haeussler et al. 2016</a>), as shown below, and the published data contains few data points with high specificity scores. Overall though, the assays showed that the higher the specificity score, the lower the off-target effects.</p> <img src="../images/crisprFig_mitScore.png"> <p>Similarly, efficiency scoring is not very accurate: guides with low scores can be efficient and vice versa. As a general rule, however, the higher the score, the less likely that a guide is very inefficient. The following histograms illustrate, for each type of score, how the share of inefficient guides drops with increasing efficiency scores: </p> <img src="../images/crisprFig_effScores.png"> <p>When reading this plot, keep in mind that both scores were evaluated on their own training data. Especially for the Moreno-Mateos score, the results are too optimistic, due to overfitting. When evaluated on independent datasets, the correlation of the prediction with other assays was around 25% lower, see <a href="#References">Haeussler et al. 2016</a>. At the time of writing, there is no independent dataset available yet to determine the Moreno-Mateos accuracy for each score percentile range.</p> <h3>Track methods</h3> <p> The entire $organism ($db) genome was scanned for the -NGG motif. Flanking 20mer guide sequences were aligned to the genome with BWA and scored with MIT Specificity scores using the command-line version of crispor.org. Non-unique guide sequences were skipped. Flanking sequences were extracted from the genome and input for Crispor efficiency scoring, available from the <a href="http://crispor.tefor.net/downloads/">Crispor downloads page</a>, which includes the Doench 2016, Moreno-Mateos 2015 and Bae 2014 algorithms, among others.</p> <p> Note that the Doench 2016 scores were updated by the Broad institute in 2017 ("Azimuth" update). As a result, earlier versions of the track show the old Doench 2016 scores and this version of the track shows new Doench 2016 scores. Old and new scores are almost identical, they are correlated to 0.99 and for more than 80% of the guides the difference is below 0.02. However, for very few guides, the difference can be bigger. In case of doubt, we recommend the new scores. Crispor.org can display both scores and many more with the "Show all scores" link.</p> <H2>Data Access</H2> <p> Positional data can be explored interactively with the <a href="../cgi-bin/hgTables?db=${db}&hgta_track=crisprAllTargets&hgta_regionType=range">Table Browser</a>. For small programmatic positional queries, the track can be accessed using our <a href="/goldenPath/help/api.html">REST API</a>. For genome-wide data or automated analysis, CRISPR genome annotations can be downloaded from <a href="http://hgdownload.soe.ucsc.edu/gbdb/$db/crisprAll/" target="_blank">our download server</a> as a bigBedFile.</p> <p> The files for this track are called <tt>crispr.bb</tt>, which lists positions and scores, and <tt>crisprDetails.tab</tt>, which has information about off-target matches. Individual regions or whole genome annotations can be obtained using our tool <tt>bigBedToBed</tt>, which can be compiled from the source code or downloaded as a pre-compiled binary for your system. Instructions for downloading source code and binaries can be found <a href="http://hgdownload.soe.ucsc.edu/downloads.html#utilities_downloads">here</a>. The tool can also be used to obtain only features within a given range, e.g.</p> <p> <tt>bigBedToBed http://hgdownload.soe.ucsc.edu/gbdb/${db}/${track}/crispr.bb -chrom=chr21 -start=0 -end=1000000 stdout</tt> </p> <h2>Credits</h2> <p> Track created by Maximilian Haeussler, with helpful input from Jean-Paul Concordet (MNHN Paris) and Alberto Stolfi (NYU). </p> <a name="References"></a> <h2>References</h2> <p> Haeussler M, Schönig K, Eckert H, Eschstruth A, Mianné J, Renaud JB, Schneider-Maunoury S, Shkumatava A, Teboul L, Kent J <em>et al</em>. <a href="https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-1012-2" target="_blank">Evaluation of off-target and on-target scoring algorithms and integration into the guide RNA selection tool CRISPOR</a>. <em>Genome Biol</em>. 2016 Jul 5;17(1):148. PMID: <a href="https://www.ncbi.nlm.nih.gov/pubmed/27380939" target="_blank">27380939</a>; PMC: <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4934014/" target="_blank">PMC4934014</a> </p> <p> Bae S, Kweon J, Kim HS, Kim JS. <a href="https://www.nature.com/nmeth/journal/v11/n7/full/nmeth.3015.html" target="_blank"> Microhomology-based choice of Cas9 nuclease target sites</a>. <em>Nat Methods</em>. 2014 Jul;11(7):705-6. PMID: <a href="https://www.ncbi.nlm.nih.gov/pubmed/24972169" target="_blank">24972169</a> </p> <p> Doench JG, Fusi N, Sullender M, Hegde M, Vaimberg EW, Donovan KF, Smith I, Tothova Z, Wilen C, Orchard R <em>et al</em>. <a href="https://www.nature.com/articles/nbt.3437" target="_blank"> Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9</a>. <em>Nat Biotechnol</em>. 2016 Feb;34(2):184-91. PMID: <a href="https://www.ncbi.nlm.nih.gov/pubmed/26780180" target="_blank">26780180</a>; PMC: <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4744125/" target="_blank">PMC4744125</a> </p> <p> Hsu PD, Scott DA, Weinstein JA, Ran FA, Konermann S, Agarwala V, Li Y, Fine EJ, Wu X, Shalem O <em>et al</em>. <a href="https://www.nature.com/nbt/journal/v31/n9/full/nbt.2647.html" target="_blank"> DNA targeting specificity of RNA-guided Cas9 nucleases</a>. <em>Nat Biotechnol</em>. 2013 Sep;31(9):827-32. PMID: <a href="https://www.ncbi.nlm.nih.gov/pubmed/23873081" target="_blank">23873081</a>; PMC: <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3969858/" target="_blank">PMC3969858</a> </p> <p> Moreno-Mateos MA, Vejnar CE, Beaudoin JD, Fernandez JP, Mis EK, Khokha MK, Giraldez AJ. <a href="https://www.nature.com/nmeth/journal/v12/n10/full/nmeth.3543.html" target="_blank"> CRISPRscan: designing highly efficient sgRNAs for CRISPR-Cas9 targeting in vivo</a>. <em>Nat Methods</em>. 2015 Oct;12(10):982-8. PMID: <a href="https://www.ncbi.nlm.nih.gov/pubmed/26322839" target="_blank">26322839</a>; PMC: <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4589495/" target="_blank">PMC4589495</a> </p>