Genomics is the study of our genetic material. In the clinic, genomic data can be used to guide diagnostic and therapeutic decision making. The concept that the effectiveness of a treatment can be determined by a patient’s genotype has a long history1. However, adoption of genome-level information in medicine only became possible with the development of next-generation sequencing technologies (NGS) and the expansion of computing infrastructure, making the “genome of 1 000 $ “a reality2. In addition, the latest advances in sequencers have also reduced the average time to sequence a human genome to as little as an hour.3. While the cost of generating genomic data is no longer prohibitive, the analysis and interpretation of the large amount of data generated by these sequencing technologies and the use of this data for clinical applications remains a challenge.

Some challenges are inherent in the technologies themselves. For example, nanopore-based sequencing technologies rely on changes in an electric current when a piece of nucleic acid passes through a membrane through a nanopore. While nanopore sequencing technology offers advantages such as long sequencing reads and direct detection of nucleic acid changes, its error rate is relatively high. However, with the advent of AI, deep learning algorithms have been developed to improve raw reading accuracies.4 and correct errors5 post-basecall.

Other challenges are inherent in the nature of the samples. Tumor samples, for example, can be very heterogeneous, harboring many somatic mutations that occur at low frequencies. This complicates the variant calling process and requires refinement to remove false positives. However, it has been shown that machine learning6 can be applied to automate the refinement step of the variant invoking cancer sequencing data, which would otherwise require manual review of aligned reads.

Interpretation of a variant can take different directions, depending on the type of the variant (such as a single nucleotide variant (SNV), insertion / deletion, etc.) and the location of the variant. For example, if an SNV is located in a coding region resulting in an amino acid change, understanding how the structure of the protein may be affected could provide insight into a disease mechanism. However, deep neural networks, which are a powerful form of AI, have recently been used to develop a method for 3D modeling proteins.7 which has demonstrated the ability to predict the structure of a protein from its genetic sequence with much higher accuracy compared to other methods. When a variant is located in a non-coding region, the variant may still exert biological consequences such as, for example, by affecting the modification of DNA or the binding of transcription factors. De novo sequence-based prediction of the effects of non-coding variants8 based on deep learning has been reported to predict certain characteristics of chromatin with high accuracy. The framework was developed to predict ab initio the effects of variants on gene expression levels9.

As mentioned above, genomic medicine involves the use of genomic data in diagnostic and therapeutic decision making. There are now many examples of rare and undiagnosed diseases diagnosedten with exome or genome sequencing. For example, in an earlier study11, exome sequencing was used in a patient with a condition similar to Crohn’s disease, but no definitive diagnosis could be made based on conventional clinical evaluation. However, exome sequencing revealed a new mutation in a gene involved in the inflammatory response and programmed cell death, but was not previously associated with Crohn’s disease. The diagnosis led to effective treatment.

More recently, an active research area consists of integrating data from omics with clinical and environmental data. Due to the complexity and amount of data involved, the use of AI tools has made it possible to determine patterns from these disparate types of data and make predictions of them. For example, checkpoint inhibitors are very effective for advanced cancer in some patients, but response rates vary from patient to patient. However, machine learning algorithms trained on data derived from whole exome sequencing, RNA-Seq, and clinical features have been used successfully to predict patient response.12 checkpoint inhibitor immunotherapy. Additionally, the integration of genomic and environmental exposure data collected via portable biosensors using machine learning methods can improve our understanding of the complexities of gene-environment interactions.13and has potential applications in health management.

Patent trends can be used to provide insight into business activity in specific areas or sectors. An analysis of AI and genomics patent filings shows that in the United States, the number of patent applications per year has more than doubled since 2015 (Figure 1)14 with a similar trend observed in Canada (Figure 2). The general trend indicates accelerated adoption of AI technologies in commercial health-related products and services.


In Canada, recent developments in patent law have the potential to provide a more favorable environment for the patenting of these technologies. Specifically, Yves Choueifaty c. Attorney General of Canada15 and resulting practice notice on patentable subject matter issued by CIPO improves the likelihood of medical diagnostic methods being patentable subject matter in Canada16. These developments may create opportunities for businesses and other organizations to generate value through patent protection, thus creating a virtuous circle for the development of additional diagnostic tools that can increase the adoption of omics and other types of big data platforms in medicine and the consumer well-being space.

As a result, in the future, we expect continued increase in innovation for AI tools that can be developed to understand disease mechanisms, discover therapeutic targets, or evaluate treatment outcomes for the benefit of. individualized patient care. For those who wish to protect these innovations to improve their ROI and create value, please do not hesitate to contact one of the members of our AI practice group.17.

This is the seventh article in our Spotlight on AIseries. You can read the first six articles here:


14. We analyzed trends in patent applications using the PatSnap® database and a search strategy adapted from WIPO Technology Trends 2019 Artificial Intelligence ( export / sites / www / tech_trends / en / artificial_intelligence / docs /techtrends_ai_methodology.pdf)

The content of this article is intended to provide a general guide on the subject. Specialist advice should be sought regarding your particular situation.

Leave a Reply

Your email address will not be published.