The prediction of gene alterations could help researchers develop effective drug treatments for cancers, says this UCD research fellow.
As a first-year student, Dr. Luis F Iglesias-Martinez was inspired by the wealth of the knowledge of his teachers.
“I remember that the conferences rejected me for how much the teachers knew,” he tells Siliconrepublic.com.
“It was they who inspired me to become a researcher.”
After completing a degree in Chemical Engineering at the National Autonomous University of Mexico, Iglesias-Martinez completed a master’s degree in biotechnology at the Technological University of Chalmers, Bepore, addressing Dublin to complete a doctorate at University College Dublin).
His doctorate developed algorithms to study patterns of gene expression in breast cancer. After this, he conducted a postdoctoral investigation that performs data analysis and develops AI methods for various conditions.
Now, he leads a group in Biology of Ireland systems, which is part of the UCD School of Medicine, where he is developing AI and automatic learning models to investigate childhood cancer genetics.
Tell us about your current research.
At the moment, we are working on a new exciting project called Magic-I, which represents the molecular and genomic interrogation for childhood-irland cancers.
This is a study in collaboration with Children’s Health Ireland, UCD, Precision Oncology Ireland, Geseq and Illumina.
Our goal is to sequence the entire genome and transcriptoma of each case of childhood cancer in Ireland over the next five years.
This is because cancer is caused by somatic genetic alterations, that is, genetic changes with which patients are not born, but that develop only in certain cells.
At the years, researchers have managed to exploit these alterations to develop cancer medicines that have improved the results of patients.
What interests us is to develop AI models that can improve the way patients are treated are treated by finding new genetic alteration patterns that boost childhood cancers.
In your opinion, why is your research important?
Fortunately, children with cancer do it much better today.
For many childhood cancers, the survival rate is high with approximately 90 % of children that survive more than five years after diagnosis.
However, there are still many children whose cans do not respond to current treatments and some of the patients who survive, do so with side effects that change their lives.
The AI models could change this, not replacing doctors, but by becoming another tool in a doctor’s arsenal in the fight against cancer, as well as radiographs and magnetic resonance machines now.
What are some of the biggest challenges or erroneous concepts that he faces as a researcher in his field?
My field lies at the intersection of AI with medicine and unfortunately both issues are plagued by erroneous concepts.
In terms of AI, many people tend to think about it as computer programs when they are actually mathematical models. Understand that they are mathematical models, I think they would help clarify their behavior.
The AI models do not think about the logical steps, but they calculate many mathematical operations with parameters that were refined from the model that tried a similar task on and again.
In medicine, and in cancer specifically, there are several erroneous concepts that I have heard many times. The ones that worry the most are the wrong concepts around which the risk of cancer increases. Unfortunately, there is a tendency to focus on things without impact on cancer risk, as if a vegetable is organic or not, while neglecting well -thought -meditated risk factors such as alcohol, obesity, smoking and human papilloma.
Do you think that public commitment to science and data has changed in recent years?
Public commitment to science has greatly increased in recent years. I think that Covid-19 pandemic pushed people to find answers to very uncertain times.
Unfortunately, social networks also cool a platform for a lot of wrong information that led people to choose not to participate in incorrect statements.
I think this has grown the legs over the years and, although it is easy to blame social networks, podcasters and influencers and poorly informed charlatans, scientists share part of the fault.
We need to make an effort to reach the public with precise information in a way that is understandable.
Do not miss the knowledge you need to succeed. Register for him Daily letterSilicon Digest Republic of Science Fiction Technology News.
