Studies of genomics make use of high throughput technology to discover and characterize genes associated with cancer and other illnesses. Genomics may be of particular value in discovering mechanisms and interventions for neurodegenerative and rare diseases in the quest for orphan drugs. To expedite risk prediction, mechanism of action and drug discovery, effectively, analytical methods, especially those that translate to clinical relevant outcomes, are highly important. In this paper, we define the term biomedical robot as a novel tool for genomic analysis in order to improve phenotype prediction, identifying disease pathogenesis and significantly defining therapeutic targets. The implementation of a biomedical robot in genomic analysis is based on the use of feature selection methods and ensemble learning techniques. Mathematically, a biomedical robot exploits the structure of the uncertainty space of any classification problem conceived as in an ill-posed optimization problem, that is, given a classifier several equivalent low scale signatures exist providing similar prediction accuracies. As an example, we applied this method to the analysis of three different expression microarrays publically available concerning Chronic Lymphocytic Leukemia, Inclusion Body Myositis/Polimyositis (IBM-PM) and Amyotrophic Lateral Sclerosis (ALS). Using these examples we showed the value of the biomedical robot concept to improve knowledge discovery and provide decision systems in order to optimize diagnosis, treatment and prognosis. The goal of the FINISTERRAE project is to leverage publically available genetic databases of rare and neurodegenerative diseases and construct a relational database with the genes and genetic pathways involved, which can be used to accelerate translational research in this domain.
[1]
Enrique J. deAndrés-Galiana,et al.
Supervised Classification by Filter Methods and Recursive Feature Elimination Predicts Risk of Radiotherapy-Related Fatigue in Patients with Prostate Cancer
,
2014,
Cancer informatics.
[2]
Alfonso Valencia,et al.
Transcriptome characterization by RNA sequencing identifies a major molecular and clinical subdivision in chronic lymphocytic leukemia
,
2014,
Genome research.
[3]
Michael J. Tompkins,et al.
On the topography of the cost functional in linear and nonlinear inverse problems
,
2012
.
[4]
G. Pinkus,et al.
Plasma cells in muscle in inclusion body myositis and polymyositis
,
2005,
Neurology.
[5]
R. Strausberg,et al.
Oncogenomics and the development of new cancer therapies
,
2004,
Nature.
[6]
S. Perrin,et al.
From transcriptome analysis to therapeutic anti-CD40L treatment in the SOD1 model of amyotrophic lateral sclerosis
,
2010,
Nature Genetics.
[7]
Juan Luis Fernández-Martínez,et al.
The effect of noise and Tikhonov's regularization in inverse problems. Part I: The linear case
,
2014
.
[8]
Juan Luis Fernández-Martínez,et al.
From Bayes to Tarantola: New insights to understand uncertainty in inverse problems☆
,
2013
.
[9]
Ana Cernea,et al.
Exploring the Uncertainty Space of Ensemble Classifiers in Face Recognition
,
2015,
Int. J. Pattern Recognit. Artif. Intell..