Riboflow: using deep learning to classify riboswitches with ~99% accuracy

Premkumar, Keshav Aditya R; Bharanikumar, Ramit; Palaniappan, Ashok

doi:10.1101/868695

Cited by 2 publications

(3 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For each sub-sequence, we extracted the new features and applied the ensemble classifier. Figure 5A shows the resulting probability of selection as a riboswitch versus the fraction of the sequence used for all (48,031) 5’UTRs; Fig. 5B shows the same result but only for the subset of 436 5’UTRs (0.91%) that were previously identified as likely riboswitches using the full length sequence; and Fig.…”

Section: Resultsmentioning

confidence: 82%

“…For the task of computational riboswitch prediction, previous methods leverage hidden Markov models (HMMER, RiboSW, Riboswitch Scanner (60; 9; 40)), covariance models and context free grammar (Infernal (42)), and sequence alignment + computational folding (Riboswitch Finder, RibEx, RNAConSLOpt (7; 1; 30)). Other more recent software such as Riboflow utilizes deep learning classifiers such as RNN-LSTM or convolutional neural networks (CNN) for their riboswitch identification (48). Computational methods available to the public up until 2018 are reviewed in Antunes et al extensively (4).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Identification of potential riboswitch elements inHomo SapiensmRNA 5’UTR sequences using Positive-Unlabeled machine learning

Raymond,

DeRoo,

Munsky

2023

Preprint

View full text Add to dashboard Cite

Riboswitches are a class of noncoding RNA structures that interact with target ligands to cause a conformational change that can then execute some regulatory purpose within the cell. Riboswitches are ubiquitous and well characterized in bacteria and prokaryotes, with additional examples also being found in fungi, plants, and yeast. To date, no purely RNA-small molecule riboswitch has been discovered inHomo Sapiens. Several analogous riboswitch-like mechanisms have been described within theH. Sapienstranslatome within the past decade, prompting the question: Is there aH. Sapiensriboswitch dependent on only small molecule ligands? In this work, we set out to train positive unlabeled machine learning classifiers on known riboswitch sequences and apply the classifiers toH. SapiensmRNA 5’UTR sequences found in the 5’UTR database, UTRdb, in the hope of identifying a set of mRNAs to investigate for riboswitch functionality. 67,683 riboswitch sequences were obtained from RNAcentral and sorted for ligand type and used as positive examples and 48,031 5’UTR sequences were used as unlabeled, unknown examples. Positive examples were sorted by lig- and, and 20 positive-unlabeled classifiers were trained on sequence and secondary structure features while withholding one or two ligand classes. Cross validation was then performed on the withheld ligand sets to obtain a validation accuracy range of 75%-99%. The joint sets of 5’UTRs identified as potential riboswitches by the 20 classifiers were then analyzed. 15333 sequences were identified as a riboswitch by one or more classifier(s) and 436 of theH. Sapiens5’UTRs were labeled as harboring potential riboswitch elements by all 20 classifiers. These 436 sequences were mapped back to the most similar riboswitches within the positive data and examined. An online database of identified and ranked 5’UTRs, their features, and their most similar matches to known riboswitches, is provided to guide future experimental efforts to identifyH. Sapiensriboswitches.Author summaryRiboswitches are an important regulatory element in bacteria that have not been described inHomo Sapiens. However, if human riboswitches exist and if they can be found, they could have vast implications on human disease. We apply positive-unlabeled machine learning to combine known riboswitch sequences withH. Sapiens5’UTR sequences and to search for potential riboswitches. We analyze our ensemble predictions for likelyH. Sapiens5’UTR riboswitches using GO analysis to determine their potential functional roles, and we rank and display our predicted sequences next to the most similar known riboswitches. We expect these analyses to be helpful to the scientific community in planning future experiments for laboratory discovery and validation.0.1Graphical Abstract

show abstract

Section: Resultsmentioning

confidence: 82%

Section: Introductionmentioning

confidence: 99%

Identification of potential riboswitch elements inHomo SapiensmRNA 5’UTR sequences using Positive-Unlabeled machine learning

Raymond,

DeRoo,

Munsky

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…We are grateful to the School of Chemical and Biotechnology, SASTRA Deemed University for infrastructure and computing support. A preliminary version of the manuscript has been released as a Pre-Print at bioRxiv (Premkumar et al, 2019).…”

Section: Acknowledgmentsmentioning

confidence: 99%

Riboflow: Using Deep Learning to Classify Riboswitches With ∼99% Accuracy

Premkumar

Bharanikumar

Palaniappan

2020

Front. Bioeng. Biotechnol.

Self Cite

View full text Add to dashboard Cite

Riboswitches are cis-regulatory genetic elements that use an aptamer to control gene expression. Specificity to cognate ligand and diversity of such ligands have expanded the functional repertoire of riboswitches to mediate mounting apt responses to sudden metabolic demands and signal changes in environmental conditions. Given their critical role in microbial life, riboswitch characterisation remains a challenging computational problem. Here we have addressed the issue with advanced deep learning frameworks, namely convolutional neural networks (CNN), and bidirectional recurrent neural networks (RNN) with Long Short-Term Memory (LSTM). Using a comprehensive dataset of 32 ligand classes and a stratified train-validate-test approach, we demonstrated the accurate performance of both the deep learning models (CNN and RNN) relative to conventional hyperparameter-optimized machine learning classifiers on all key performance metrics, including the ROC curve analysis. In particular, the bidirectional LSTM RNN emerged as the best-performing learning method for identifying the ligand-specificity of riboswitches with an accuracy >0.99 and macro-averaged F-score of 0.96. An additional attraction is that the deep learning models do not require prior feature engineering. A dynamic update functionality is built into the models to factor for the constant discovery of new riboswitches, and extend the predictive modeling to new classes. Our work would enable the design of genetic circuits with custom-tuned riboswitch aptamers that would effect precise translational control in synthetic biology. The associated software is available as an open-source Python package and standalone resource for use in genome annotation, synthetic biology, and biotechnology workflows.

show abstract

Riboflow: using deep learning to classify riboswitches with ~99% accuracy

Cited by 2 publications

References 59 publications

Identification of potential riboswitch elements inHomo SapiensmRNA 5’UTR sequences using Positive-Unlabeled machine learning

Identification of potential riboswitch elements inHomo SapiensmRNA 5’UTR sequences using Positive-Unlabeled machine learning

Riboflow: Using Deep Learning to Classify Riboswitches With ∼99% Accuracy

Contact Info

Product

Resources

About