A primer to frequent itemset mining for bioinformatics

Naulaerts, Stefan; Meysman, Pieter; Bittremieux, Wout; Vu, Trung Nghia; Berghe, Wim Vanden; Goethals, Bart; Laukens, Kris

doi:10.1093/bib/bbt074

Cited by 108 publications

(107 citation statements)

References 79 publications

Supporting

Mentioning

107

Contrasting

Order By: Relevance

“…The archetypical example of frequent itemset mining is the discovery of products that are frequently purchased together from mining large numbers of supermarket basket transactions. The study of such associations, for example the observation that beer and chips are frequently bought together, is a computationally non-trivial problem for which various algorithms have been developed, as reviewed before [18]. Despite the explosive number of possible patterns with growing datasets, frequent itemset mining techniques are available to efficiently extract all possible patterns even from large and complex databases.…”

Section: Resultsmentioning

confidence: 99%

Unravelling associations between unassigned mass spectrometry peaks with frequent itemset mining techniques

Mrzic

Valkenborg

et al. 2014

Proteome Sci

View full text Add to dashboard Cite

BackgroundMass spectrometry-based proteomics experiments generate spectra that are rich in information. Often only a fraction of this information is used for peptide/protein identification, whereas a significant proportion of the peaks in a spectrum remain unexplained. In this paper we explore how a specific class of data mining techniques termed “frequent itemset mining” can be employed to discover patterns in the unassigned data, and how such patterns can help us interpret the origin of the unexpected/unexplained peaks.ResultsFirst a model is proposed that describes the origin of the observed peaks in a mass spectrum. For this purpose we use the classical correlative database search algorithm. Peaks that support a positive identification of the spectrum are termed explained peaks. Next, frequent itemset mining techniques are introduced to infer which unexplained peaks are associated in a spectrum. The method is validated on two types of experimental proteomic data. First, peptide mass fingerprint data is analyzed to explain the unassigned peaks in a full scan mass spectrum. Interestingly, a large numbers of experimental spectra reveals several highly frequent unexplained masses, and pattern mining on these frequent masses demonstrates that subsets of these peaks frequently co-occur. Further evaluation shows that several of these co-occurring peaks indeed have a known common origin, and other patterns are promising hypothesis generators for further analysis. Second, the proposed methodology is validated on tandem mass spectrometral data using a public spectral library, where associations within the mass differences of unassigned peaks and peptide modifications are explored. The investigation of the found patterns illustrates that meaningful patterns can be discovered that can be explained by features of the employed technology and found modifications.ConclusionsThis simple approach offers opportunities to monitor accumulating unexplained mass spectrometry data for emerging new patterns, with possible applications for the development of mass exclusion lists, for the refinement of quality control strategies and for a further interpretation of unexplained spectral peaks in mass spectrometry and tandem mass spectrometry.Electronic supplementary materialThe online version of this article (doi:10.1186/s12953-014-0054-1) contains supplementary material, which is available to authorized users.

show abstract

Section: Resultsmentioning

confidence: 99%

Unravelling associations between unassigned mass spectrometry peaks with frequent itemset mining techniques

Mrzic

Valkenborg

et al. 2014

Proteome Sci

View full text Add to dashboard Cite

show abstract

“…Table 4 describes the size of the used data sets, the number of transactions and the number of items in each of the transactions. We have compared the performance of the Pe-ARM with a set of well known association rule mining algorithms (BSO-ARM [27], ACO R [13], SA [40], G3PARM [11], ARMBGSA [25]). The parameters used by these algorithms are the optimal values proposed by the authors.…”

Section: Evaluation With Standard Data Setsmentioning

confidence: 99%

Penguins Search Optimisation Algorithm for Association Rules Mining

Gheraibia¹

2016

CIT. J. comput. inf. technol.

View full text Add to dashboard Cite

Association Rules Mining (ARM) is one of the most popular and well-known approaches for the decisionmaking process. All classic exhaustive ARM algorithms are time consuming and generate a very large number of association rules, even the recent proposed meta-heuristics based methods generate a small number of high quality rules but with high overlapping. To deal with this issue, we propose a new ARM approach based on penguins search optimisation algorithm (Pe-ARM for short). Moreover, an efficient measure is incorporated into the main process to evaluate the amount of overlapping among the generated rules. The proposed approach also ensures a good diversification over the whole solutions space. To demonstrate the effectiveness of the proposed approach, several experiments have been carried out on different data sets and specifically on the biological ones. The results reveal that the proposed approach outperforms the well known meta-heuristics ARM algorithms in both execution time and solution quality.

show abstract

“…Although the application of association rules (AR) [7] could appear as a trivial task, it is less popular than other analytical techniques, such as statistics methods. Literature reports different uses of AR mining from annotated data that are usually tailored to some specific aims as reported by Naulaerts et al [8]. For instance, Faria et al [9] proposed association rules to support GO curators by evaluating the annotation consistency in order to avoid possible inconsistent or redundant annotations.…”

Section: Introductionmentioning

confidence: 99%

Extracting Cross-Ontology Weighted Association Rules from Gene Ontology Annotations

Agapito

Milano

Guzzi

et al. 2016

IEEE/ACM Trans. Comput. Biol. and Bioinf.

View full text Add to dashboard Cite

Gene Ontology (GO) is a structured repository of concepts (GO Terms) that are associated to one or more gene products through a process referred to as annotation. The analysis of annotated data is an important opportunity for bioinformatics. There are different approaches of analysis, among those, the use of association rules (AR) which provides useful knowledge, discovering biologically relevant associations between terms of GO, not previously known. In a previous work, we introduced GO-WAR (Gene Ontology-based Weighted Association Rules), a methodology for extracting weighted association rules from ontology-based annotated datasets. We here adapt the GO-WAR algorithm to mine cross-ontology association rules, i.e., rules that involve GO terms present in the three sub-ontologies of GO. We conduct a deep performance evaluation of GO-WAR by mining publicly available GO annotated datasets, showing how GO-WAR outperforms current state of the art approaches.

show abstract

A primer to frequent itemset mining for bioinformatics

Cited by 108 publications

References 79 publications

Unravelling associations between unassigned mass spectrometry peaks with frequent itemset mining techniques

Unravelling associations between unassigned mass spectrometry peaks with frequent itemset mining techniques

Penguins Search Optimisation Algorithm for Association Rules Mining

Extracting Cross-Ontology Weighted Association Rules from Gene Ontology Annotations

Contact Info

Product

Resources

About