Designing viral diagnostics with model-based optimization

Metsky, Hayden C.; Welch, Nicole L.; Pillai, Priya P.; Haradhvala, Nicholas J.; Rumker, Laurie; Mantena, Sreekar; Zhang, Yibin B.; Yang, David; Ackerman, Cheri M.; Weller, Juliane; Blainey, Paul C.; Myhrvold, Cameron; Mitzenmacher, Michael; Sabeti, Pardis C.

doi:10.1101/2020.11.28.401877

Cited by 10 publications

(16 citation statements)

References 110 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consistent with our original correlation analysis (Figure 2A), the CNN and GBT models had a clear preference for an alternating stretch of guanines, cytosines and guanines (G 15-18 C 19-22 G 23-24 ) in this core region (Figure 5B and 5D). This unique core motif was not found for Cas13a when we performed a correlational analysis of available datasets (Abudayyeh et al, 2017; Metsky et al, 2021) (Figure S8). Indeed, no consistent sequence motif or core region emerged across the Cas13a datasets analyzed, which could be due to intrinsic enzymatic properties of Cas13a or limitations in the size of available datasets.…”

Section: Resultsmentioning

confidence: 89%

“…Importantly, analysis of base preference at the individual nucleotide level only obscured this motif, underscoring the importance of motif-level approaches to model interpretation such as TF-MoDISco used here - the first time to our knowledge such a motif-level approach has been applied to CRISPR guide activity prediction. Analysis of available Cas13a datasets (Abudayyeh et al, 2017; Metsky et al, 2021) did not indicate a similar motif, suggesting that it may be unique to Cas13d.…”

Section: Discussionmentioning

confidence: 95%

“…We analyzed three Cas13a guide efficiency datasets: 1) the Luciferase knockdown dataset containing 186 LwaCas13a guides for Gaussia luciferase (Gluc) and 93 guides for Cypridina Luciferase (Cluc) (Abudayyeh et al, 2017); 2) the endogenous gene knockdown dataset containing 93 LwaCas13a guides for each of KRAS, PPIB and MALAT1 (Abudayyeh et al, 2017); and 3) the ADAPT dataset containing 85 perfect match LwaCas13a guides for virus detection (Metsky et al, 2021). We calculated the Pearson correlation between each nucleotide at each position with guide efficiency to evaluate the sequence contribution.…”

Section: Methodsmentioning

confidence: 99%

“…279 LwaCas13a guides for KRAS, PPIB and MALAT1 were analyzed and the Pearson correlation coefficient for each positional nucleotide with guide efficiency is shown. C. Correlation of each nucleotide with guide efficiency at each guide position in the LwaCas13a ADAPT dataset (Metsky et al, 2021). 85 perfect match guides from LwaCas13a ADAPT data were analyzed and the Pearson correlation coefficient for each positional nucleotide with guide efficiency is shown.…”

Section: Supplementary Figurementioning

confidence: 99%

See 3 more Smart Citations

Deep learning and CRISPR-Cas13d ortholog discovery for optimized RNA targeting

Wei

Lotfy

Faizi

et al. 2021

Preprint

View full text Add to dashboard Cite

Transcriptome engineering requires flexible RNA-targeting technologies that can perturb mammalian transcripts in a robust and scalable manner. CRISPR systems that natively target RNA molecules, such as Cas13 enzymes, are enabling rapid progress in the investigation of RNA biology and advancement of RNA therapeutics. Here, we sought to develop a Cas13 platform for high-throughput phenotypic screening and elucidate the design principles underpinning its RNA targeting efficiency. We employed the RfxCas13d (CasRx) system in a positive selection screen by tiling 55 known essential genes with single nucleotide resolution. Leveraging this dataset of over 127,000 guide RNAs, we systematically compared a series of linear regression and machine learning algorithms to train a convolutional neural network (CNN) model that is able to robustly predict guide RNA performance based on guide sequence alone. We further incorporated secondary features including secondary structure, free energy, target site position, and target isoform percent. To evaluate model performance, we conducted orthogonal screens via cell surface protein knockdown. The final CNN model is able to predict highly effective guide RNAs (gRNAs) within each transcript with >90% accuracy in this independent test set. To provide user interpretability, we evaluate feature contributions using both integrated gradients and SHapley Additive exPlanations (SHAP). We identify a specific sequence motif at guide position 15-24 along with selected secondary features to be predictive of highly efficient guides. Taken together, we derive Cas13d guide design rules from large-scale screen data, release a guide design tool (http://rnatargeting.org) to advance the RNA targeting toolbox, and describe a path for systematic development of deep learning models to predict CRISPR activity.

show abstract

Section: Resultsmentioning

confidence: 89%

Section: Discussionmentioning

confidence: 95%

Section: Methodsmentioning

confidence: 99%

Section: Supplementary Figurementioning

confidence: 99%

See 2 more Smart Citations

Deep learning and CRISPR-Cas13d ortholog discovery for optimized RNA targeting

Wei

Lotfy

Faizi

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…2A). For this reason, the selection of "good" crRNAs that support efficient Cas13 activity is critical for bulk Cas13-based molecular diagnostics 15 , though how different crRNAs affect the activity of Cas13 is not well understood 16 . We therefore tested crRNAs 11 and 12 in our droplet assay and compared it to the activity of crRNA 4.…”

Section: Crrna/target Rna Combinations Govern Cas13a Reaction Kineticsmentioning

confidence: 99%

Sensitive and multiplexed RNA detection with Cas13 droplets and kinetic barcoding

Son

Lyden

Shu

et al. 2021

Preprint

View full text Add to dashboard Cite

SUMMARYRapid and sensitive quantification of RNA is critical for detecting infectious diseases and identifying disease biomarkers. Recent direct detection assays based on CRISPR-Cas13a1–4 avoid reverse transcription and DNA amplification required of gold-standard PCR assays5, but these assays have not yet achieved the sensitivity of PCR and are not easily multiplexed to detect multiple viruses or variants. Here we show that Cas13a acting on single target RNAs loaded into droplets exhibits stochastic nuclease activity that can be used to enable sensitive, rapid, and multiplexed virus quantification. Using SARS-CoV-2 RNA as the target and combinations of CRISPR RNA (crRNA) that recognize different parts of the viral genome, we demonstrate that reactions confined to small volumes can rapidly achieve PCR-level sensitivity. By tracking nuclease activity within individual droplets over time, we find that Cas13a exhibits rich kinetic behavior that depends on both the target RNA and crRNA. We demonstrate that these kinetic signatures can be harnessed to differentiate between different human coronavirus species as well as SARS-CoV-2 variants within a single droplet. The combination of high sensitivity, short reaction times, and multiplexing makes this droplet-based Cas13a assay with kinetic barcoding a promising strategy for direct RNA identification and quantification.

show abstract