CollHaps: A Heuristic Approach to Haplotype Inference by Parsimony

Abstract

Haplotype data play a relevant role in several genetic studies, e.g., mapping of complex disease genes, drug design, and evolutionary studies on populations. However, the experimental determination of haplotypes is expensive and time-consuming. This motivates the increasing interest in techniques for inferring haplotype data from genotypes, which can instead be obtained quickly and economically. Several such techniques are based on the maximum parsimony principle, which has been justified by both experimental results and theoretical arguments. However, the problem of haplotype inference by parsimony was shown to be NP-hard, thus limiting the applicability of exact parsimony-based techniques to relatively small data sets. In this paper, we introduce collapse rule, a generalization of the well-known Clark's rule, and describe a new heuristic algorithm for haplotype inference (implemented in a program called CollHaps), based on parsimony and the iterative application of collapse rules. The performance of CollHaps is tested on several data sets. The experiments show that CollHaps enables the user to process large data sets obtaining very "parsimonious" solutions in short processing times. They also show a correlation, especially for large data sets, between parsimony and correct reconstruction, supporting the validity of the parsimony principle to produce accurate solutions.

Haplotype inference based on sparse dictionary selection

“…All methods have the same genotype error rate except for CollHaps, which has a higher error rate. This poor result of CollHaps was expected, as is stressed in [6] the importance of using data sets with a large number of individuals in order to achieve a good performance by this method.…”

“…The first one consists of synthetic data generated using the coalescence theory. The second one corresponds to the Angiotensin Converting Enzyme (ACE) data set that is considered to be a data set with a number of subjects not large enough in comparison with the number of distinct haplotypes [6].…”

“…In particular, we present the results of applying PHASE [7] and its fast version fastPHASE [9], Gerbil [8] and CollHaps [6] to the same data sets. Synthetic Data: Haplotypes are simulated and randomly paired to form genotypes of short sequence data (between 5 and 30 SNPs in each haplotype).…”

“…The set of estimated haplotypes in this case depends on the order in which the genotypes are given. A generalization to Clark's rule was presented in [6], where another heuristic called CollHaps is used to find the haplotypes.…”

A GRASP-based approach for the Pure Parsimony Haplotype Inference problem

“…Such a sequence has the same allele on a site as the two (unknown) haplotype sequences if they agree on that site and otherwise is regarded as ambiguous. Combined with the fact that the determination of haplotype sequences via experiments is usually highly costly and time consuming [3], [4] it is therefore not surprising that the problem of inferring for a set of genotype sequences a set of, in a welldefined sense, explaining haplotype sequences has received a considerable amount of attention in the literature. For example, cast as a statistical problem it has led to powerful approaches such as PHASE [5] which have turned out to be particularly useful in case the amount of variability in a set of genotype sequences is high (see [6], [7] for more on this and [8] for a recent overview on statistical approaches).…”