Artificial Intelligence and Cancer (Part 2): Big Data and Precision Oncology

“Big Data” and “Artificial Intelligence” are some of the buzzwords most people use nowadays. However, in the medical community, it is proving to be a largely advantageous tool for multiple fields. Doctors have always taken great notice of medical information coming from substances to records. It is a tool that helps them assess and treat their patients. Incorporating larger sets of data into medical information provides greater insight and perspective into a subject. In terms of oncology, it can help in several ways such as treatment and prediction. To better understand the hot topic of artificial intelligence and big data, we will look into its specific applications and recent achievements.

Artificial Intelligence and Big Data

Simply put, big data refers to a very large set of data amounting to terabytes and even zettabytes. To put it into perspective, a terabyte is 1000 times larger than the usual gigabyte, while a zettabyte is 1,000,000,000,000 gigabytes. Statisticians often use advanced analytics to make sense of these large sets of data.

If you are to pair it with artificial intelligence, then the processing of this data scales in leaps and bounds. Common industry applications of this include drastic improvements in decision-making and creating models for future outcomes and predictions. From these applications, we can see how they can add significant help to the medical field. 

Healthcare has long been an expertise that involves data storage. The introduction of big data and artificial intelligence (AI) has helped in many medical applications. These include integration of health records for personalized medicine, management of medical records, and assessment of doctor’s findings through natural language processing. Specifically for oncology, there are a lot of databases that can help in researching cancer. 

Artificial Intelligence and Big Data for Precision Oncology

There have already been a lot of different studies detailing the contribution of AI into cancer research. In this 2019 review article on the overview of precision medicine using AI, researchers note the application of deep learning to the data from medical image records. It has seen outstanding results in the diagnosis of metastatic breast cancer, melanoma, and retinal diseases, even outperforming pathologists and dermatologists. 

While oncology refers to the general research and therapy of tumor-related conditions, precision oncology digs deeper into the genetic and molecular study of these tumors. Researchers in the field of gene editing are seeing the use of big data and AI through CRISPR, a precise laboratory tool used to manipulate DNA. These applications are extremely applicable to one of the fastest-growing fields of clinical research in multi-omics. 

Multi-omics refers to an approach in data analysis encompassing “omic” fields of biology. These fields include the genome, proteome, transcriptome, epigenome, metabolome, and microbiome. All of these integrate with one another, and their big data sets provide incredible specificity of a subject, such as patient-specific information. Readily-available sources of reference data help this field greatly throughout its analysis. These are some of the databases:

  • Cancer Genome Atlas: includes detailed information on cancer patients.
  • DrugBank: includes detailed information on drugs, nutraceutical structures, and metabolite structures.
  • PubChem: a chemistry database on all commercially-available compounds and other synthesizable compounds.
  • Protein Data Bank: three-dimensional crystal structure database of proteins and their ligands.

Big Data Applications in Cancer Research

From the following benefits of AI and big data exhibited above, we can see how these two computer technologies advance the medical field. If we are to explore further, we shall discuss the specific applications of these technologies into precision oncology. Here are some of these applications. 

Targeted Drug Discovery

While the aforementioned databases are of help to many fields of medicine, these are specifically used in drug development and discovery. In particular, this 2019 review article shows how AI and big data used these databases to develop a novel targeted drug.

Targeted drugs are important to cancer treatment because tumor cells are normally surrounded by healthy cells. Doctors want to get rid of the tumor cells, not the healthy cells around it. To address these problems, researchers look to lead compounds with remarkable properties of Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET). There are several ways they used AI and big data for targeted drug discovery:

  • Virtual screening, selection, and scoring to discover targeted compounds using computer-aided drug design 
  • Repurposing of old drugs through AI reverse docking of several data sets.
  • Using deep learning models to boost and predict a drug’s ADMET properties. 
  • Utilizing 3D structural alignment for the specificity of therapeutic targets between human and animal models.

Next Generation Sequencing

As cancer is largely a genetic disease, one computer-derived aid that has excellent potential in oncology is next-generation sequencing (NGS). It is one of the recent developments in DNA sequence analysis. NGS boasts a performance of millions to billions of simultaneous and separate sequencing, allowing for a complete human genome analysis in a few days. NGS provides other capabilities, including:

  • Sequence alterations in an entire genome
  • Sequence an entire transcriptome or transcribed RNA
  • Identify genetic translocations
  • Show expression levels of a gene

Cancer Prediction

When we refer to big data, we mean huge numbers. These numbers showed up in a 2020 study on trying to predict factors for cancer survival using AI. The researchers used data from the Surveillance, Epidemiology, and End Results from 1975 to 2016. This amounted to 257,880 cases of oral cancer in 40 years. 

From this study, the researchers found that the primary factors for cancer survival were age at diagnosis, the primary site of cancer, tumor size, and the year of diagnosis. The average length of survival for all patients was 60.35 months after diagnosis. 

We can see from all these studies how AI and big data can help medical practitioners in medical decisions and treatment. Furthermore, all these applications in multiple fields help in developing an emerging treatment called precision medicine. This personalized treatment takes into account the factors around genetics, environment, lifestyle choices, and the individual drug-response of the person. This treatment will surely help in terms of cancer survival. 

Click here for our blog Disclaimer.