Apply artificial intelligence to visualize genetic data of melanoma patients

Apply artificial intelligence to visualize genetic data of melanoma patients

Microscopic image of a melanoma, one of the most serious cancers that is represented on the skin. (Image: F. Descubre)

Researchers from the Department of Computer Languages ​​and Sciences of the University of Malaga, together with the Oncology Unit of the Regional and Virgen de la Victoria Hospitals of Málaga   and the Institute of Biomedical Research (IBIMA) of Málaga (Spain), have developed a computer application that allows to obtain clinical data of melanoma patients simply and quickly.

Based on genetic databases of patients, they have designed a new tool that integrates this existing information and improves the visualization of data to physicians. With this, it is possible the receipt of patterns of those affected by this skin disease and compare the evolution of the disease in different cases.

The proposed solution is based, first of all, on the filtering of genes from patients. This allows selecting those that represent the state in which the disease is or the evolution of it. Once these more characteristic genes are filtered, a visualization is offered through a heat map, which allows us to differentiate by color those genes that behave similarly. In this way, the expert can observe them better and establish coincidences in patients who have had a similar medical history. These similarities serve as an initial clue to establish indicators of the evolution of those affected.

The second type of analysis implemented in the application is a graph that shows the genes that are related by expressing themselves similarly over a time series.

According to experts, this work mainly manages to perform a more effective initial analysis of the available data and make it available to clinical researchers. As indicated by the researcher, Ismael Navas Delgado to the Fundación Discover: “There was already a specific data analysis software, but only aimed at geneticists or bioinformatics experts. That is why we have developed a simple tool, which allows the clinical researcher to undertake a first exploration of the expression data, to detect patterns or genes of interest in the clinical trial ”.

In the study entitled ‘VIGLA-M: visual gene expression data analytics’, published in the journal BMC Bioinformatics, has used as an starting point an anonymous database of patients participating in this clinical trial of which they have stored several samples to throughout the treatment they follow. In each sample, 770 genes are analyzed, from which the level of expression is obtained and, with it, the response to the treatment they are obtaining. Thus, the new application is able to reduce the samples to about 6 more representative genes of the patients, which is what allows the simplicity of data for later clinical review.

Melanoma is currently positioned as one of the most serious cancers that is represented on the skin. One of its causes is exposure to ultraviolet rays from solar radiation, which means that adequate control of sun exposure can prevent this disease. It is a very harmful type of cancer in its cases of complication and metastasis, so it is important for clinical professionals to have as much data as possible about the evolution of each patient and their response to treatments. Thus, once again, the ‘big data’ technology provides the oncologist with a database on the history of patient response, in order to be able to establish a comparison with other cases and thus make a thorough clinical study.

As Ismael Navas points out: “We have started with a reduced sample for the initial tests, which we will increase progressively, and that has served as a basis for perfecting the use of the tool.” That is why, currently scientists continue to work in the same line. On the one hand, the tool has been perfected for use by clinical staff so that it is already usable by them. On the other hand, we have worked on an algorithm that achieves better visual results.

The study is funded by projects of the Ministry of Science, Innovation and Universities, a project of the Andalusian Research, Development and Innovation Plan of the Junta de Andalucía, by a Post-Doctoral Grant of ‘Talent Collection for Research’ of the Own Plan from the University of Málaga, the teaching project ‘Open Big Data Analytics for eHealth’ of the Andalusia TECH International Campus of Excellence and by the Progress and Health Foundation of the Junta de Andalucía.

(Source: F. Descubre y Amazings® / NCYT® | (Noticiasdelaciencia.com / Amazings.com)).

Close Menu
×

Cart