Learning Comprehensible Classification Rules From Gene Expression Data Using Genetic Programming and Biological Ontologies

Goertzel, Ben; Coelho, Lúcio; Pennachin, Cassio; Goertzel, Izabela Freire; Queiroz, Murilo; Prosdocimi, Francisco; Lobo, Francisco Pereira

doi:10.1142/9789812774118_0081

Cited by 2 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SVM results for aging brains are taken from [13]. These essentially used a meta-tasking framework where algorithm parameters (e.g., the SVM kernel) were varied across a large number of runs (on portions training data) to determine the best configuration, which was then applied on all of the training data.…”

Section: Classification Resultsmentioning

confidence: 99%

“…A number of approaches to handling large feature spaces have been proposed in conjunction with program evolution for learn classifiers, most commonly either preliminary runs to identify promising features [19,1], or inexpensive preprocessing heuristics [18,13]. We follow the later, selecting simply the most-differentiating features to use in all experiments (similar to the signal-to-noise approach found to perform well in [18]).…”

Section: Concerning Microarray Gene Expression Datamentioning

confidence: 99%

“…Past studies have applied GP to learn classifiers that operate directly on gene expression levels [16,18,1,19,13], which are manipulated with arithmetic and other operators. However, microarray data are very noisy [20,39]; we conjecture that, by and large, considering the actual gene expression levels without restriction will increase the danger of overfitting and make the learning of good models more difficult (by expanding the hypothesis space is a non-productive manner).…”

Section: Concerning Microarray Gene Expression Datamentioning

confidence: 99%

“…However, it has been found that GP is sometimes preferable to SVMs and other methodologies in this context because of the transparency of the models it produces. Studying the GP models themselves, can give valuable insights into exactly how the classification is being performed, which can lead to biological insights [12,13,14].…”

mentioning

confidence: 99%

“…Here we report results on two of these datasets in detail. One of the datasets poses the problem of distinguishing between types of lymphoma based on gene expression data [34,36,13]; the other poses the problem of classifying between young and old human brains [23,13,14]. In our discussion, we aim to illustrate not only the quantitative merit of the results, but also their apparent biological meaningfulness.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Understanding microarray data through applying competent program evolution

Looks

Goertzel²,

Coelho³

et al. 2007

Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

A number of researchers have used genetic programming (GP) to analyze gene expression microarray data; here it is asked whether the use of an alternate program evolution technique, MOSES (meta-optimizing semantic evolutionary search), can improve upon GP's results in this domain. Based on our results so far, the answer appears to be a resounding yes. We first consider standard supervised classification for two microarray datasets: one involving aging vs. young brains, and the other involving lymphoma types. On these datasets, MOSES is shown to learn classifiers with test accuracies comparable to those obtained by SVMs, and significantly superior to those of GP with a similar configuration (and even more superior to those of GP run under a more standard configuration). Furthermore, unlike SVMs, the classification rules produced by MOSES are still simple and comprehensible to humans -often strikingly simple. Finally, biological interpretation is presented regarding the "important genes" used most often in the ensemble of MOSES models learned; we show that many of the genes identified in this way reflect knowledge available in the literature, and others constitute plausible hypotheses.

show abstract

Section: Classification Resultsmentioning

confidence: 99%

Section: Concerning Microarray Gene Expression Datamentioning

confidence: 99%

Section: Concerning Microarray Gene Expression Datamentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Understanding microarray data through applying competent program evolution

Looks

Goertzel²,

Coelho³

et al. 2007

Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

show abstract

Identifying the Genes and Genetic Interrelationships Underlying the Impact of Calorie Restriction on Maximum Lifespan: An Artificial Intelligence-Based Approach

Goertzel¹,

Pennachin²,

Mudado³

et al. 2008

Rejuvenation Research

View full text Add to dashboard Cite

Novel artificial intelligence methodologies were applied to analyze gene expression microarray data gathered from mice under a calorie restriction (CR) regimen. The data were gathered from three previously published mouse studies; these datasets were merged together into a single composite dataset for the purpose of conducting a broader-based analysis. The result was a list of genes that are important for the impact of CR on lifespan, not necessarily in terms of their individual actions but in terms of their interactions with other genes. Furthermore, a map of gene interrelationships was provided, suggesting which intergene interactions are most important for the effect of CR on life extension. In particular our analysis showed that the genes Mrpl12, Uqcrh, and Snip1 play central roles regarding the effects of CR on life extension, interacting with many other genes (which the analysis enumerates) in carrying out their roles. This is the first time that the genes Snip1 and Mrpl12 have been identified in the context of aging. In a follow-up analysis aimed at validating these results, the analytic process was rerun with a fourth dataset included, yielding largely comparable results. Broadly, the biological interpretation of these analytical results suggests that the effects of CR on life extension are due to multiple factors, including factors identified in prior theories of aging, such as the hormesis, development, cellular, and free radical theories.

show abstract

Learning Comprehensible Classification Rules From Gene Expression Data Using Genetic Programming and Biological Ontologies

Cited by 2 publications

References 0 publications

Understanding microarray data through applying competent program evolution

Understanding microarray data through applying competent program evolution

Identifying the Genes and Genetic Interrelationships Underlying the Impact of Calorie Restriction on Maximum Lifespan: An Artificial Intelligence-Based Approach

Contact Info

Product

Resources

About