In a new study presented at the meeting of the American Society of Haematology (ASH) conference, the researchers suggested that forging AI (Artificial intelligence) with genomics will push precision medicine forward in blood cancer and lead to new insights and discoveries.   

In 2017, approximately 62,130 people are expected to be diagnosed with leukaemia in the U.S. The numbers are staggering, and scientists do not fully understand the causes of the disease. It is essential to understand how large the genome of a single individual is.

To help unlock some of the secrets hidden in the human genome, IBM and the Munich Leukaemia Laboratory (MLL) analyzed genomic data that include whole genome sequencing (WGS) and transcriptome sequencing (RNA-Seq) in at least 5,000 cases of patients with leukaemia.

Blood disorders are one area of research in which genomics holds considerable promise. Every year, clinicians, pathologist, researchers of haematology from around the world descend at the meeting of the American Society of Haematology (ASH).

MLL and IBM presented at the conference, suggesting the melding of AI with genomics and bioinformatics to accelerate precision medicine in hematologic malignancies. The results suggested that forging AI with genomics will not just propel precision medicine forward, but also lead to new insights and discoveries.

AI has penetrated the lives through smartphones, IoTs and driverless cars, so it is only natural to invite AI into the labs of ASH participants. Our initial research has been to demonstrate that agnostic genomic AI approaches can be validated, meaning we can prove these systems work by testing it against things we already know.

As a first step, we've been able to prove that these AI approaches can identify genes already known to be implicated in certain sub-types of blood cancer (such as gene JAK2 in ET and PMF subtypes, ASXL1 in aCML and CMML, KIT in SM-AHN and SM subtypes of blood cancer).

Moving beyond validating AI approaches, the researchers are identifying indicators of disease subtypes from non-gene parts of the genome. The promise of AI combined with genomic and bioinformatics could offer clinicians new insights and understanding into devastating diseases like leukaemia.