DNA and RNA have been compared to “instruction manuals” containing the information needed for living “machines” to operate. But while electronic machines like computers and robots are designed from the ground up to serve a specific purpose, biological organisms are governed by a much messier, more complex set of functions that lack the predictability of binary code. Inventing new solutions to biological problems requires teasing apart seemingly intractable variables – a task that is daunting to even the most intrepid human brains.

Two teams of scientists from the Wyss Institute at Harvard University and the Massachusetts Institute of Technology have devised pathways around this roadblock by going beyond human brains; they developed a set of machine learning algorithms that can analyze reams of RNA-based “toehold” sequences and predict which ones will be most effective at sensing and responding to a desired target sequence. As reported in two papers published concurrently today in Nature Communications, the algorithms could be generalizable to other problems in synthetic biology as well, and could accelerate the development of biotechnology tools to improve science and medicine and help save lives.

“These achievements are exciting because they mark the starting point of our ability to ask better questions about the fundamental principles of RNA folding, which we need to know in order to achieve meaningful discoveries and build useful biological technologies,” said Luis Soenksen, Ph.D., a Postdoctoral Fellow at the Wyss Institute and Venture Builder at MIT’s Jameel Clinic who is a co-first author of the first of the two papers.

Getting ahold of toehold switches

The collaboration between data scientists from the Wyss Institute’s Predictive BioAnalytics Initiative and synthetic biologists in Wyss Core Faculty member Jim Collins’ lab at MIT was created to apply the computational power of machine learning, neural networks, and other algorithmic architectures to complex problems in biology that have so far defied resolution. As a proving ground for their approach, the two teams focused on a specific class of engineered RNA molecules: toehold switches, which are folded into a hairpin-like shape in their “off” state. When a complementary RNA strand binds to a “trigger” sequence trailing from one end of the hairpin, the toehold switch unfolds into its “on” state and exposes sequences that were previously hidden within the hairpin, allowing ribosomes to bind to and translate a downstream gene into protein molecules. This precise control over the expression of genes in response to the presence of a given molecule makes toehold switches very powerful components for sensing substances in the environment, detecting disease, and other purposes.

However, many toehold switches do not work very well when tested experimentally, even though they have been engineered to produce a desired output in response to a given input based on known RNA folding rules. Recognizing this problem, the teams decided to use machine learning to analyze a large volume of toehold switch sequences and use insights from that analysis to more accurately predict which toeholds reliably perform their intended tasks, which would allow researchers to quickly identify high-quality toeholds for various experiments.

The first hurdle they faced was that there was no dataset of toehold switch sequences large enough for deep learning techniques to analyze effectively. The authors took it upon themselves to generate a dataset that would be useful to train such models. “We designed and synthesized a massive library of toehold switches, nearly 100,000 in total, by systematically sampling short trigger regions along the entire genomes of 23 viruses and 906 human transcription factors,”  said Alex Garruss, a Harvard graduate student working at the Wyss Institute who is a co-first author of the first paper. “The unprecedented scale of this dataset enables the use of advanced machine learning techniques for identifying and understanding useful switches for immediate downstream applications and future design.”

Armed with enough data, the teams first employed tools traditionally used for analyzing synthetic RNA molecules to see if they could accurately predict the behavior of toehold switches now that there were manifold more examples available. However, none of the methods they tried – including mechanistic modeling based on thermodynamics and physical features – were able to predict with sufficient accuracy which toeholds functioned better.

A picture is worth a thousand base pairs

The researchers then explored various machine learning techniques to see if they could create models with better predictive abilities. The authors of the first paper decided to analyze toehold switches not as sequences of bases, but rather as two-dimensional “images” of base-pair possibilities. “We know the baseline rules for how an RNA molecule’s base pairs bond with each other, but molecules are wiggly – they never have a single perfect shape, but rather a probability of different shapes they could be in,” said Nicolaas Angenent-Mari, a MIT graduate student working at the Wyss Institute and co-first author of the first paper. “Computer vision algorithms have become very good at analyzing images, so we created a picture-like representation of all the possible folding states of each toehold switch, and trained a machine learning algorithm on those pictures so it could recognize the subtle patterns indicating whether a given picture would be a good or a bad toehold.”

Another benefit of their visually-based approach is that the team was able to “see” which parts of a toehold switch sequence the algorithm “paid attention” to the most when determining whether a given sequence was “good” or “bad.” They named this interpretation approach Visualizing Secondary Structure Saliency Maps, or VIS4Map, and applied it to their entire toehold switch dataset. VIS4Map successfully identified physical elements of the toehold switches that influenced their performance, and allowed the researchers to conclude that toeholds with more potentially competing internal structures were “leakier” and thus of lower quality than those with fewer such structures, providing insight into RNA folding mechanisms that had not been discovered using traditional analysis techniques.

“Being able to understand and explain why certain tools work or don’t work has been a secondary goal within the artificial intelligence community for some time, but interpretability needs to be at the forefront of our concerns when studying biology because the underlying reasons for those systems’ behaviors often cannot be intuited,” said Jim Collins, Ph.D., the senior author of the first paper. “Meaningful discoveries and disruptions are the result of deep understanding of how nature works, and this project demonstrates that machine learning, when properly designed and applied, can greatly enhance our ability to gain important insights about biological systems.” Collins is also the Termeer Professor of Medical Engineering and Science at MIT.

Now you’re speaking my language

While the first team analyzed toehold switch sequences as 2D images to predict their quality, the second team created two different deep learning architectures that approached the challenge using orthogonal techniques. They then went beyond predicting toehold quality and used their models to optimize and redesign poorly performing toehold switches for different purposes, which they report in the second paper.

The first model, based on a convolutional neural network (CNN) and multi-layer perceptron (MLP), treats toehold sequences as 1D images, or lines of nucleotide bases, and identifies patterns of bases and potential interactions between those bases to predict good and bad toeholds. The team used this model to create an optimization method called STORM (Sequence-based Toehold Optimization and Redesign Model), which allows for complete redesign of a toehold sequence from the ground up. This “blank slate” tool is optimal for generating novel toehold switches to perform a specific function as part of a synthetic genetic circuit, enabling the creation of complex biological tools.

“The really cool part about STORM and the model underlying it is that after seeding it with input data from the first paper, we were able to fine-tune the model with only 168 samples and use the improved model to optimize toehold switches. That calls into question the prevailing assumption that you need to generate massive datasets every time you want to apply a machine learning algorithm to a new problem, and suggests that deep learning is potentially more applicable for synthetic biologists than we thought,” said co-first author Jackie Valeri, a graduate student at MIT and the Wyss Institute.

The second model is based on natural language processing (NLP), and treats each toehold sequence as a “phrase” consisting of patterns of “words,” eventually learning how certain words are put together to make a coherent phrase. “I like to think of each toehold switch as a haiku poem: like a haiku, it’s a very specific arrangement of phrases within its parent language – in this case, RNA. We are essentially training this model to learn how to write a good haiku by feeding it lots and lots of examples,” said co-first author Pradeep Ramesh, Ph.D., a Visiting Postdoctoral Fellow at the Wyss Institute and Machine Learning Scientist at Sherlock Biosciences.

Ramesh and his co-authors integrated this NLP-based model with the CNN-based model to create NuSpeak (Nucleic Acid Speech), an optimization approach that allowed them to redesign the last 9 nucleotides of a given toehold switch while keeping the remaining 21 nucleotides intact. This technique allows for the creation of toeholds that are designed to detect the presence of specific pathogenic RNA sequences, and could be used to develop new diagnostic tests.

The team experimentally validated both of these platforms by optimizing toehold switches designed to sense fragments from the SARS-CoV-2 viral genome. NuSpeak improved the sensors’ performances by an average of 160%, while STORM created better versions of four “bad” SARS-CoV-2 viral RNA sensors whose performances improved by up to 28 times.

“A real benefit of the STORM and NuSpeak platforms is that they enable you to rapidly design and optimize synthetic biology components, as we showed with the development of toehold sensors for a COVID-19 diagnostic,” said co-first author Katie Collins, an undergraduate MIT student at the Wyss Institute who worked with MIT Associate Professor Timothy Lu, M.D., Ph.D., a corresponding author of the second paper.

“The data-driven approaches enabled by machine learning open the door to really valuable synergies between computer science and synthetic biology, and we’re just beginning to scratch the surface,” said Diogo Camacho, Ph.D., a corresponding author of the second paper who is a Senior Bioinformatics Scientist and co-lead of the Predictive BioAnalytics Initiative at the Wyss Institute. “Perhaps the most important aspect of the tools we developed in these papers is that they are generalizable to other types of RNA-based sequences such as inducible promoters and naturally occurring riboswitches, and therefore can be applied to a wide range of problems and opportunities in biotechnology and medicine.”

Additional authors of the papers include Wyss Core Faculty member and Professor of Genetics at HMS George Church, Ph.D.; and Wyss and MIT Graduate Students Miguel Alcantar and Bianca Lepe.

“Artificial intelligence is wave that is just beginning to impact science and industry, and has incredible potential for helping to solve intractable problems. The breakthroughs described in these studies demonstrate the power of melding computation with synthetic biology at the bench to develop new and more powerful bioinspired technologies, in addition to leading to new insights into fundamental mechanisms of biological control,” said Don Ingber, M.D., Ph.D., the Wyss Institute’s Founding Director. Ingber is also the Judah Folkman Professor of Vascular Biology at Harvard Medical School and the Vascular Biology Program at Boston Children’s Hospital, as well as Professor of Bioengineering at Harvard’s John A. Paulson School of Engineering and Applied Sciences.

The post Deep learning Gets a “Toehold” on Synthetic Biology appeared first on Artificial Intelligence.

A new study led by researchers at Linköping University demonstrates how an artificial neural network (ANN) can reveal large amounts of gene expression data, and it can lead to the discovery of groups of disease-related genes. The study was published in Nature Communications, and the scientists want the method to be applied within precision medicine and individualized treatment. 

Scientists are currently developing maps of biological networks that are based on how different proteins or genes interact with each other. The new study involves the use of artificial intelligence (AI) in order to find out if biological networks can be discovered through the use of deep learning. Artificial neural networks, which are trained by experimental data in the process of deep learning, are able to find patterns within massive amounts of complex data. Because of this, they are often used in applications such as image recognition. Even with its seemingly enormous potential, the use of this machine learning method has been limited within biological research. 

Sanjiv Dwivedi is a postdoc in the Department of Physics, Chemistry and Biology (IFM) at Linköping University.

“We have for the first time used deep learning to find disease-related genes. This is a very powerful method in the analysis of huge amounts of biological information, or ‘big data’,” says Dwivedi.

The scientists relied on a large database with information regarding the expression patterns of 20,000 genes in a large number of people. The artificial neural network was not told which gene expression patterns were from people with diseases, or which ones were from healthy individuals. The AI model was then trained to find patterns of gene expression.

One of the mysteries surrounding machine learning is that it is currently impossible to see how an artificial neural network gets to its final result. It is only possible to see the information that goes in and the information that is produced, but everything that happens in-between consists of several layers of mathematically processed information. These inner workings of an artificial neural network are not yet able to be deciphered. The scientists wanted to know if there were any similarities between the designs of the neural network and the familiar biological networks. 

Mike Gustafsson is a senior lecturer at IFM and leads the study. 

“When we analysed our neural network, it turned out that the first hidden layer represented to a large extent interactions between various proteins. Deeper in the model, in contrast, on the third level, we found groups of different cell types. It’s extremely interesting that this type of biologically relevant grouping is automatically produced, given that our network has started from unclassified gene expression data,” says Gustafsson.

The scientists then wanted to know if their model of gene expression was capable of being used to determine which gene expression patterns are associated with disease and which are normal. They were able to confirm that the model can discover relative patterns that agree with biological mechanisms in the body. Another discovery was that the artificial neural network could possibly discover brand new patterns since it was trained with unclassified data. The researchers will now investigate previously unknown patterns and whether they are relevant within biology. 

“We believe that the key to progress in the field is to understand the neural network. This can teach us new things about biological contexts, such as diseases in which many factors interact. And we believe that our method gives models that are easier to generalise and that can be used for many different types of biological information,” says Gustafsson.

Through collaborations with medical researchers, Gustafsson hopes to apply the method in precision medicine. This could help determine which specific types of medicine patients should receive.

The study was financially supported by the Swedish Foundation for Strategic Research (SSF) and the Swedish Research Council.

The post Deep Learning Used to Find Disease-Related Genes appeared first on Artificial Intelligence.

Machine learning enables computers to assist humans in analyzing data from giant, advanced information sets. one amongst the advanced information is biology and genomic information that has to analyze a varied set of functions mechanically by the computers. These machine learning strategies will offer additional help for creating this information for any usage like cistron prediction, organic phenomenon, cistron metaphysics, cistron finding, cistron has written material and etc. the aim of this study is to explore some machine learning applications and algorithms to genetic and genomic information.

In genetic science, machine learning will be used to learn however to extract the location and structure of varied genes,  to establish restrictive parts,  to characteristic non-coding polymer genes,  to predicting cistron operate, to predicting polymer secondary structure.

To annotate a large type of ordination sequencing parts we will use machine learning strategies. Generally, if we will compile an inventory of sequence parts of a given kind, then we will most likely train a machine learning methodology to acknowledge those parts, then models will be combined on with logic concerning their relative locations.

Many application areas that are associated with machine learning such as cistron sequence, cistron expression, macromolecule structure, cistron restrictive networks, microarrays. Before attempting to resolve the problems within the on top of the space of genetic science we tend to use a machine learning algorithm to train the system to classify the cistron information and produce a model.  Some trendy techniques to handle the cistron information will be used.

There are such a lot of machine learning algorithms is used for identification,  prediction,  selection, recognization,  and conjointly in the classification of polymer sequences. For the identification of cistron in the polymer sequence, the neural network primarily based multi-classifier will be used. the organic phenomenon is known by promoters. To predicting, the situation of the promoter neural network has been used. For predicting the promoters within the cistrons and evaluating the performance of the gene the synthetic neural network classifier has been used. In an ordination analysis,  the very important half is to learning to polymer sequence pattern recognization.

Machine learning will take as computer file generated by alternative genomic assays, like microarray or polymer. Sequence expression information, transcription issue, binding chip sequence information, etc. Another example of ordination information is cistron expression information will be used to learn to distinguish between completely different unwellness phenotypes and within the method, to spot potentially valuable unwellness biomarkers. we will conjointly use machine learning to assign annotations to cistrons these quite annotations maximum taken from the gene metaphysics assignment terms.

Future applications of machine learning in genomics:

1. As genomic information comparatively giant in size thus machine learning approaches will create that to simply analyze and create the items as simplified.

2. Gene sequencing is terribly straightforward to analyze solely by exploitation machine learning strategies. The sequence of the varied genes should have labels thus exploitation supervised learning algorithms simply gets the sequence accurately, Deep genetic science most the machine learning algorithms are won’t to for cistron sequencing.

3. Gene written material one amongst the most analysis space within the next cistronration for analyzing the genes and finding the precise matches in genes and changes the gene sequence per the means its need to target. particularly cistron written material with machine learning will scale back the time, price and energy required to spot the targeted sequence.

4. Pharmacogenomics field provides an additional advantage for initiating the customized drugs that are the drug that is given to the patient for a specific unwellness that ought to adapt to the genetic makeup of the individual patient. characteristic of the dose of drug machine learning is able to do a mass challenge in the future.

5. New Born screening tools will create use of machine learning approaches in characteristic the metabolism defects

The post Machine learning applications in genomics and genetics appeared first on Artificial Intelligence.