skip to main content

This content will become publicly available on August 13, 2023

Title: acCRISPR: An activity-correction method for improving the accuracy of CRISPR screens
High throughput CRISPR screens are revolutionizing the way scientists unravel the genetic underpinnings of novel and evolved phenotypes. One of the critical challenges in accurately assessing screening outcomes is accounting for the variability in sgRNA cutting efficiency. Poorly active guides targeting genes essential to screening conditions obscure the growth defects that are expected from disrupting them. Here, we develop acCRISPR, an end-to-end pipeline that identifies essential genes in pooled CRISPR screens using sgRNA read counts obtained from next-generation sequencing. acCRISPR uses experimentally determined cutting efficiencies for each guide in the library to provide an activity correction to the screening outcomes, thus determining the fitness effect of disrupted genes. This is accomplished by calculating an optimization metric that quantifies the tradeoff between guide activity and library coverage, which is maximized to accurately classify genes essential to screening conditions. CRISPR-Cas9 and -Cas12a screens were carried out in the non-conventional oleaginous yeast Yarrowia lipolytica to determine a high-confidence set of essential genes for growth under glucose, a common carbon source used for the industrial production of oleochemicals. acCRISPR was also used in gain-and loss-of-function screens under high salt and low pH conditions to identify known and novel genes that were related to stress more » tolerance. Collectively, this work presents an experimental-computational framework for CRISPR-based functional genomics studies that may be expanded to other non-conventional organisms of interest. « less
; ; ; ; ; ; ;
Award ID(s):
Publication Date:
Journal Name:
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Genome-wide functional genetic screens have been successful in discovering genotype-phenotype relationships and in engineering new phenotypes. While broadly applied in mammalian cell lines and inE. coli, use in non-conventional microorganisms has been limited, in part, due to the inability to accurately design high activity CRISPR guides in such species. Here, we develop an experimental-computational approach to sgRNA design that is specific to an organism of choice, in this case the oleaginous yeastYarrowia lipolytica. A negative selection screen in the absence of non-homologous end-joining, the dominant DNA repair mechanism, was used to generate single guide RNA (sgRNA) activity profiles for both SpCas9 and LbCas12a. This genome-wide data served as input to a deep learning algorithm, DeepGuide, that is able to accurately predict guide activity. DeepGuide uses unsupervised learning to obtain a compressed representation of the genome, followed by supervised learning to map sgRNA sequence, genomic context, and epigenetic features with guide activity. Experimental validation, both genome-wide and with a subset of selected genes, confirms DeepGuide’s ability to accurately predict high activity sgRNAs. DeepGuide provides an organism specific predictor of CRISPR guide activity that with retraining could be applied to other fungal species, prokaryotes, and other non-conventional organisms.

  2. Extracellular electron transfer (EET) – the process by which microorganisms transfer electrons across their membrane(s) to/from solid-phase materials – has implications for a wide range of biogeochemically important processes in marine environments. Though EET is thought to play an important role in the oxidation of inorganic minerals by lithotrophic organisms, the mechanisms involved in the oxidation of solid particles are poorly understood. To explore the genetic basis of oxidative EET, we utilized genomic analyses and transposon insertion mutagenesis screens (Tn-seq) in the metabolically flexible, lithotrophic Alphaproteobacterium Thioclava electrotropha ElOx9 T . The finished genome of this strain is 4.3 MB, and consists of 4,139 predicted ORFs, 54 contain heme binding motifs, and 33 of those 54 are predicted to localize to the cell envelope or have unknown localizations. To begin to understand the genetic basis of oxidative EET in ElOx9 T , we constructed a transposon mutant library in semi-rich media which was comprised of >91,000 individual mutants encompassing >69,000 unique TA dinucleotide insertion sites. The library was subjected to heterotrophic growth on minimal media with acetate and autotrophic oxidative EET conditions on indium tin oxide coated glass electrodes poised at –278 mV vs. SHE or un-poised in an openmore »circuit condition. We identified 528 genes classified as essential under these growth conditions. With respect to electrochemical conditions, 25 genes were essential under oxidative EET conditions, and 29 genes were essential in both the open circuit control and oxidative EET conditions. Though many of the genes identified under electrochemical conditions are predicted to be localized in the cytoplasm and lack heme binding motifs and/or homology to known EET proteins, we identified several hypothetical proteins and poorly characterized oxidoreductases that implicate a novel mechanism(s) for EET that warrants further study. Our results provide a starting point to explore the genetic basis of novel oxidative EET in this marine sediment microbe.« less
  3. In mammalian cells, nutrients and growth factors signal through an array of upstream proteins to regulate the mTORC1 growth control pathway. Because the full complement of these proteins has not been systematically identified, we developed a FACS-based CRISPR-Cas9 genetic screening strategy to pinpoint genes that regulate mTORC1 activity. Along with almost all known positive components of the mTORC1 pathway, we identified many genes that impact mTORC1 activity, includingDCAF7,CSNK2B,SRSF2,IRS4,CCDC43, andHSD17B10. Using the genome-wide screening data, we generated a focused sublibrary containing single guide RNAs (sgRNAs) targeting hundreds of genes and carried out epistasis screens in cells lacking nutrient- and stress-responsive mTORC1 modulators, including GATOR1, AMPK, GCN2, and ATF4. From these data, we pinpointed mitochondrial function as a particularly important input into mTORC1 signaling. While it is well appreciated that mitochondria signal to mTORC1, the mechanisms are not completely clear. We find that the kinases AMPK and HRI signal, with varying kinetics, mitochondrial distress to mTORC1, and that HRI acts through the ATF4-dependent up-regulation of both Sestrin2 and Redd1. Loss of both AMPK and HRI is sufficient to render mTORC1 signaling largely resistant to mitochondrial dysfunction induced by the ATP synthase inhibitor oligomycin as well as the electron transport chain inhibitors piericidinmore »and antimycin. Taken together, our data reveal a catalog of genes that impact the mTORC1 pathway and clarify the multifaceted ways in which mTORC1 senses mitochondrial dysfunction.

    « less
  4. Abstract

    Cell-based transcriptional reporters are invaluable in high-throughput compound and CRISPR screens for identifying compounds or genes that can impact a pathway of interest. However, many transcriptional reporters have weak activities and transient responses. This can result in overlooking therapeutic targets and compounds that are difficult to detect, necessitating the resource-consuming process of running multiple screens at various timepoints. Here, we present RADAR, a digitizer circuit for amplifying reporter activity and retaining memory of pathway activation. Reporting on the AP-1 pathway, our circuit identifies compounds with known activity against PKC-related pathways and shows an enhanced dynamic range with improved sensitivity compared to a classical reporter in compound screens. In the first genome-wide pooled CRISPR screen for the AP-1 pathway, RADAR identifies canonical genes from the MAPK and PKC pathways, as well as non-canonical regulators. Thus, our scalable system highlights the benefit and versatility of using genetic circuits in large-scale cell-based screening.

  5. Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)
    The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should producemore »identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster. Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive. These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations. The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity. GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient. In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc. To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise. We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues.« less