Large-scale DNA-based phenotypic recording and deep learning enable highly accurate sequence-function mapping

Abstract

10Predicting quantitative effects of gene regulatory elements (GREs) on gene expression is a longstanding 11 challenge in biology. Machine learning models for gene expression prediction may be able to address 12 this challenge, but they require experimental datasets that link large numbers of GREs to their 13 quantitative effect. However, current methods to generate such datasets experimentally are either 14 restricted to specific applications or limited by their technical complexity and error-proneness. Here we 15 introduce DNA-based phenotypic recording as a widely applicable and practical approach to generate 16 very large datasets linking GREs to quantitative functional readouts of high precision, temporal 17 resolution, and dynamic range, solely relying on sequencing. This is enabled by a novel DNA 18 architecture comprising a site-specific recombinase, a GRE that controls recombinase expression, and a 19 DNA substrate modifiable by the recombinase. Both GRE sequence and substrate state can be 20 determined in a single sequencing read, and the frequency of modified substrates amongst constructs 21 harbouring the same GRE is a quantitative, internally normalized readout of this GRE's effect on 22 recombinase expression. Using next-generation sequencing, the quantitative expression effect of 23 extremely large GRE sets can be assessed in parallel. As a proof of principle, we apply this approach to 24 record translation kinetics of more than 300,000 bacterial ribosome binding sites (RBSs), collecting over 25 2.7 million sequence-function pairs in a single experiment. Further, we generalize from these large-scale 26Recent progress in DNA sequencing and synthesis has facilitated reading and (re-)writing of the genetic 33 makeup of biological systems on a massive scale 1,2 . Despite this progress, the relationship between a 34 genetic sequence and its functional properties is poorly understood, and thus the question "what to write" 35 remains largely unanswered 3,4 . Since the number of possible sequences scales exponentially with their 36 length, the theoretical sequence space cannot be exhaustively explored by experiments, even for small 37GREs 5-7 . Therefore, innovative high-throughput (HTP) approaches are required that allow to collect a 38 quantitative functional readout for large numbers of genetic sequences 7,8 . At the same time, novel 39 methods are required that identify statistical patterns and dependencies in the resulting datasets to 40 generate models that accurately predict the properties of untested sequences. Deep learning maximizes 41 the benefit of data collection at large scale owing to its ability to capture complex, nonlinear 42 dependencies and to its computational scalability 9 , which led to several successful applications in 43 computational biology, from genomics to proteomics 10-15 . These methods promise to be able to model 44 sequence-function dependencies with minimal prior assumptions, provided that large experimental 45 training datasets that link sequence to quantitative measure ...

A deep learning approach to programmable RNA switches

“…One significant challenge of using deep learning to predict biological function is the inherent difficulty in understanding learned patterns in a way that helps researchers to elucidate biological mechanisms underlying model predictions. Recent work has been developed to visualize sequence features by mapping learned convolutional filters to biologically relevant sequence motifs 45,46 . Additional methods have been established to address how models link biological theory, including alternative network architectures 47 , and the use of saliency maps 48,49 , which reveal the regions of input that deep-learning models weigh most heavily and therefore pay the most attention to when making predictions.…”

The Ongoing Quest to Crack the Genetic Code for Protein Production

“…A recent, innovative study that used machine learning focused on predicting the influence of different 5 0 UTR sequences in E. coli (Hö llerer et al, 2020). The study developed an innovative reporter system, based on a recombinase protein, to quantify the expression from a large library of randomized 5 0 UTR sequences (Figure 4).…”

“…The latter approach was demonstrated for the expression of a randomized 5 0 UTR library mediating expression of a recombinase that flips a nearby DNA modification site. In the same single-sequencing read, the 5 0 UTR variant can be identified and whether the site was flipped or not, allowing high-quality, large-scale data on expression levels (Hö llerer et al, 2020). Analysis of generated large-scale data is typically performed by multiple regression analysis and, recently, by machine-learning algorithms.…”

Application of combinatorial optimization strategies in synthetic biology

“…sequencing a genomic region multiple times, is an NGS approach that can be applied to track the genetic heterogeneity in a library of cell isolates. Höllerer et al introduced a wildly applicable DNA-based phenotypic recording approach to generate huge datasets linking regulators to quantitative functional readouts of high precision, only relying on sequencing short tag DNA elements 77 . The technique implements a site-specific recombinase, a regulator that controls recombinase expression, and a DNA substrate modifiable by the recombinase.…”

“…Overall, results achieved from combinatorial optimization and integration of these data into available computational predictive models, can provide better understanding of the whole cellular system 95 . Höllerer et al, implemented next-generation sequencing to assess the quantitative expression effect of extremely large sets of RBSs 77 . They expanded from these large-scale datasets using a novel deep learning approach that combines ensembling and uncertainty modelling to predict the function of untested RBSs with high accuracy.…”