DPhil the Future: AI is Revolutionising Disease Detection and Gene Editing with CRISPR

DPhil the Future: AI is Revolutionising Disease Detection and Gene Editing with CRISPR by DPhil student Shruti Chakraborty.

The dual challenge of early disease detection and implementing effective treatment protocols continues to be a critical healthcare priority. A remarkable convergence of technologies - combining CRISPR's (Clustered Regularly Interspaced Short Palindromic Repeats) precision with AI’s analytical power, is transforming the future of both molecular diagnostics and genetic medicine.   

The Evolution of Molecular Detection 

Scientists have developed various methods to detect disease-causing organisms through their genetic material. PCR or Polymerase-Chain-Reaction, developed in the 1980s, revolutionised molecular diagnostics by amplifying DNA to detectable levels. However, its dependence on sophisticated laboratories and trained personnel led to the development of simpler approaches. 

These newer methods, called isothermal nucleic acid amplification techniques (iNAATs), can operate at constant temperatures, making them more practical for widespread use. Additionally, nanoparticle-based technologies emerged, using gold and silver particles to enable visual detection of pathogens. 

The Challenge of Pathogenic Variants 

As viruses and bacteria evolve, they create variants that can be difficult to distinguish from each other. During epidemics and pandemics, rapid and accurate identification of these variants becomes crucial for effective treatment and disease containment. Traditional diagnostic methods like PCR, iNAATs, and nanoparticle-based tests have long been used for genotyping and mutation detection. However, CRISPR-based paper biosensors have transformed pathogen detection with greater simplicity and accessibility. By combining CRISPR's Cas complex with iNAATs' isothermal amplification capabilities, more efficient alternatives to conventional testing methods have been derived. 

CRISPR: A Dual Revolution in Medicine 

CRISPR is a natural defence system that bacteria use to protect themselves from viruses. Scientists have adapted this system for two important medical purposes: editing genes and detecting diseases. 

In gene editing, CRISPR works like a precise targeting system. It uses a guide molecule (guide RNA) to lead an enzyme (Cas) to a specific location in DNA. When it reaches its target, the enzyme makes a precise cut in the DNA. The cell then repairs this cut, allowing scientists to either remove unwanted genes or replace them with beneficial ones. 

For disease detection, scientists use a different feature of CRISPR. When certain CRISPR enzymes find their target (like a virus's genetic material: DNA/RNA), they not only cut that target but also start cutting other nearby genetic materials in its vicinity. Scientists have exploited this collateral effect for nucleic acid (DNA/RNA) detection. This makes CRISPR a powerful tool for identifying pathogens and its variants quickly and accurately. 

This dual ability - to both edit genes and detect diseases - makes CRISPR one of the most versatile tools in modern medicine. Its precision and adaptability have opened new possibilities for treating genetic diseases and improving disease diagnosis. 

In a groundbreaking demonstration of CRISPR gene therapy, scientists successfully modified blood stem cells from patients with severe blood disorders (beta-thalassemia and sickle cell disease) to correct their haemoglobin production. The modified cells, when returned to patients, produced healthy haemoglobin - freeing them from blood transfusions and disease symptoms for over a year. More details can be found in: https://www.nejm.org/doi/pdf/10.1056/NEJMoa2031054  

Also, several innovative CRISPR-based diagnostic platforms have been produced. For example, DETECTR specialises in DNA virus detection, particularly HPV, while SHERLOCK enables sensitive detection of RNA viruses including SARS-CoV-2. These platforms offer significant advantages: rapid results, simple operation, and accessibility, particularly valuable in resource-limited settings and during epidemics or pandemics. 

Artificial Intelligence: Enhancing CRISPR's Precision 

A critical challenge in CRISPR applications is ensuring precise targeting. The system must effectively locate and modify intended genetic sequences (on-target activity) while avoiding similar but unintended sequences (off-target effects). Traditional experimental methods for validating these activities are labour-intensive, time-consuming, and costly. 

AI has been instrumental in designing optimal guide RNAs with high on-target efficiency and minimal off-target effects. By analysing large datasets of previous experiments, AI systems learn to predict which guide RNAs will work best and where they might cause unwanted effects (quantify off-target activities), significantly reducing the time and cost of developing new genetic therapies. This replaces months of arduous laboratory testing with rapid computational predictions, making CRISPR technology more practical for medical applications. 

Several deep-learning and machine-learning studies using sophisticated neural networks (e.g. CNN, RNN, GNN, Transformers), have been employed to assess potential on-target efficiencies and off-target activities. There are tools built with AI (e.g. CRISPOR, inDelphi, DeepSpCas9) designed to enhance CRISPR's precision and safety in medical applications. These methods and available tools significantly enhance the safety and reliability of CRISPR applications in medicine. More details can be found in: https://link.springer.com/protocol/10.1007/978-1-0716-4079-1_17#   

Advanced Genetic Technologies and AI 

This field continues to advance with more sophisticated genetic modification techniques. Base editing enables precise single-nucleotide modifications in DNA, while prime editing allows for targeted insertion of genetic sequences. AI platforms like BE-HIVE and PRIDICT are instrumental in optimising these advanced applications, analysing multiple factors to predict editing outcomes and enhance success rates. 

Looking Forward 

The integration of AI with CRISPR technology represents a significant advancement in medical science. These developments are making precise disease detection and genetic treatments increasingly accessible and reliable. While technical challenges persist, the synergy between CRISPR and AI continues to expand the possibilities in molecular medicine, promising improved diagnostic and therapeutic options for various diseases. 

The rapid advancement of genetic medicine, particularly in CRISPR-based gene editing, is poised for further acceleration with the integration of sophisticated computational approaches. Powerful AI methodologies like Generative AI, LLMs, Graph Neural Networks, Geometric Deep Learning and Topological Data Analysis offer the potential to unlock deeper understandings of CRISPR complex structures and functions. Although still nascent in their application to CRISPR research, these advanced computational tools, proven effective in drug discovery and protein engineering, hold immense potential for revolutionising our understanding of the CRISPR-Cas complex. This enhanced comprehension promises to significantly refine and augment gene editing capabilities. The increasing adoption of these tools heralds a future where CRISPR technology achieves unprecedented precision and efficacy, unlocking new frontiers in genetic medicine. 

This writing is summarised from my recent publications: 

Chakraborty, S.S., Ray Dutta, J., Ganesan, R., Minary, P. (2025). The Evolution of Nucleic Acid–Based Diagnosis Methods from the (pre-)CRISPR to CRISPR era and the Associated Machine/Deep Learning Approaches in Relevant RNA Design. In: Churkin, A., Barash, D. (eds) RNA Design. Methods in Molecular Biology, vol 2847. Humana, New York, NY. https://doi.org/10.1007/978-1-0716-4079-1_17