BACKGROUND: Variation in mitochondrial DNA (mtDNA) identified by genotyping microarrays or by sequencing only the hypervariable regions of the genome may be insufficient to reliably assign mitochondrial genomes to phylogenetic lineages or haplogroups. This lack of resolution can limit functional and clinical interpretation of a substantial body of existing mtDNA data. To address this limitation, we developed and evaluated a large, curated reference alignment of complete mtDNA sequences as part of a pipeline for imputing missing mtDNA single nucleotide variants (mtSNVs). We call our reference alignment and pipeline MitoImpute. RESULTS: We aligned the sequences of 36,960 complete human mitochondrial genomes downloaded from GenBank, filtered and controlled for quality. These sequences were reformatted for use in imputation software, IMPUTE2. We assessed the imputation accuracy of MitoImpute by measuring haplogroup and genotype concordance in data from the 1000 Genomes Project and the Alzheimer's Disease Neuroimaging Initiative (ADNI). The mean improvement of haplogroup assignment in the 1000 Genomes samples was 42.7% (Matthew's correlation coefficient = 0.64). In the ADNI cohort, we imputed missing single nucleotide variants. CONCLUSION: These results show that our reference alignment and panel can be used to impute missing mtSNVs in existing data obtained from using microarrays, thereby broadening the scope of functional and clinical investigation of mtDNA. This improvement may be particularly useful in studies where participants have been recruited over time and mtDNA data obtained using different methods, enabling better integration of early data collected using less accurate methods with more recent sequence data.

Microsatellites are short (2-6bp) DNA sequences repeated in tandem, which make up approximately 3% of the human genome [ 1 ]. These loci are prone to frequent mutations and high polymorphism with the estimated mutation rates of 10 - 10 events per locus per generation, orders of magnitude higher than other parts of the genome [ 2 ]. Dozens of neurological and developmental disorders have been attributed to microsatellite expansions [ 3 ]. Microsatellites have also been implicated in a range of functions such as DNA replication and repair, chromatin organisation and regulation of gene expression [ 4 ]. Traditionally, microsatellite variation has been measured using capillary gel electrophoresis [ 5 ]. In addition to being time-consuming, and expensive, this method fails to reveal the full complexity at these loci because it does not directly sequence the fragment but only measure the number of bases in the repeat. Next-generation sequencing has the potential to address these problems. However, determining microsatellite lengths using next-generation sequencing data is difficult. In particular, polymerase slippage during PCR amplification introduces stutter noise. A small number of software tools have been written to genotype simple microsatellites in next-generation sequencing data [ 6 – 8 ], however they fail to address the issues of SNPs and compound repeats, and in some cases provide only approximate genotypes. We have begun to develop a microsatellite genotyping algorithm that addresses these issues, providing high accuracy as well as more detailed analysis of microsatellite loci. We have validated it using high depth amplicon sequencing data of microsatellites near the AVPR1A gene. We found high concordance between our algorithm and repeat lengths obtained by electrophoresis, manual inspection and Mendelian inheritance (Table 1 ). By subsampling the reads, we found that our model is accurate to within one repeat unit down to coverages that we would expect in standard exome sequencing (Figure 1 ). Additionally, we detected polymorphic single nucleotide changes within some microsatellites. Genotyping accuracy at the (AC) n promoter locus as a function of the number of reads spanning the microsatellite . 20 to 3000 reads were sampled with replacement from those spanning the microsatellite. This was done 1000 times for each depth. A shows the portion of genotypes that were exactly correct, B shows the proportion of genotypes that were correct to within one repeat unit. The algorithm was approximately 95% correct at calling the exact same genotype on high depth sequencing data. When it did call a genotype incorrectly, the genotype was only one repeat unit different. The algorithm can perform at approximately 90% accuracy to within one repeat unit with as few as 20 informative reads and reaches almost 100% accuracy to within one repeat unit with 100 or more informative reads. Future work will include expanding the algorithm to genotype compound microsatellites and further validation and comparison with other algorithms will be performed on whole genome data sets.

BACKGROUND: Continuous-time Markov models allow flexible, parametrically succinct descriptions of sequence divergence. Non-reversible forms of these models are more biologically realistic but are challenging to develop. The instantaneous rate matrices defined for these models are typically transformed into substitution probability matrices using a matrix exponentiation algorithm that employs eigendecomposition, but this algorithm has characteristic vulnerabilities that lead to significant errors when a rate matrix possesses certain 'pathological' properties. Here we tested whether pathological rate matrices exist in nature, and consider the suitability of different algorithms to their computation. RESULTS: We used concatenated protein coding gene alignments from microbial genomes, primate genomes and independent intron alignments from primate genomes. The Taylor series expansion and eigendecomposition matrix exponentiation algorithms were compared to the less widely employed, but more robust, Padé with scaling and squaring algorithm for nucleotide, dinucleotide, codon and trinucleotide rate matrices. Pathological dinucleotide and trinucleotide matrices were evident in the microbial data set, affecting the eigendecomposition and Taylor algorithms respectively. Even using a conservative estimate of matrix error (occurrence of an invalid probability), both Taylor and eigendecomposition algorithms exhibited substantial error rates: ~100% of all exonic trinucleotide matrices were pathological to the Taylor algorithm while ~10% of codon positions 1 and 2 dinucleotide matrices and intronic trinucleotide matrices, and ~30% of codon matrices were pathological to eigendecomposition. The majority of Taylor algorithm errors derived from occurrence of multiple unobserved states. A small number of negative probabilities were detected from the Padé algorithm on trinucleotide matrices that were attributable to machine precision. Although the Padé algorithm does not facilitate caching of intermediate results, it was up to 3x faster than eigendecomposition on the same matrices. CONCLUSION: Development of robust software for computing non-reversible dinucleotide, codon and higher evolutionary models requires implementation of the Padé with scaling and squaring algorithm.

UNLABELLED: Most phylogenetic methods assume that the sequences evolved under homogeneous, stationary and reversible conditions. Compositional heterogeneity in data intended for studies of phylogeny suggests that the data did not evolve under these conditions. SeqVis, a Java application for analysis of nucleotide content, reads sequence alignments in several formats and plots the nucleotide content in a tetrahedron. Once plotted, outliers can be identified, thus allowing for decisions on the applicability of the data for phylogenetic analysis. AVAILABILITY: http://www.bio.usyd.edu.au/jermiin/programs.htm.

MOTIVATION: Maximum-likelihood analysis of nucleotide and amino acid sequences is a powerful approach for inferring phylogenetic relationships and for comparing evolutionary hypotheses. Because it is a computationally demanding and time-consuming process, most algorithms explore only a minute portion of tree-space, with the emphasis on finding the most likely tree while ignoring the less likely, but not significantly worse, trees. However, when such trees exist, it is equally important to identify them to give due consideration to the phylogenetic uncertainty. Consequently, it is necessary to change the focus of these algorithms such that near optimal trees are also identified. RESULTS: This paper presents the Advanced Stepwise Addition Algorithm for exploring tree-space and two algorithms for generating all binary trees on a set of sequences. The Advanced Stepwise Addition Algorithm has been implemented in TrExML, a phylogenetic program for maximum-likelihood analysis of nucleotide sequences. TrExML is shown to be more effective at finding near optimal trees than a similar program, fastDNAml, implying that TrExML offers a better approach to account for phylogenetic uncertainty than has previously been possible. A program, TreeGen, is also described; it generates binary trees on a set of sequences allowing for extensive exploration of tree-space using other programs. AVAILABILITY: TreeGen, TrExML, and the sequence data used to test the programs are available from the following two WWW sites: http://whitetail.bemidji.msus. edu/trexml/and http://jcsmr.anu.edu.au/dmm/humgen.+ ++html.

A general algorithm is presented for identifying sets of positions in multiple sequence alignments that best characterize an a priori partitioning such as those determined by inhibition studies or other experimental techniques. The algorithm explores combinations of polymorphic columns in the alignment and evaluates how well these sites reflect the original input partition. Partitions across the polymorphic columns are derived using a tree building procedure with conventional amino acid substitution matrices. Elucidation of those amino acids which govern the biochemical behaviour of a protein with a given substrate or inhibitor can provide insights towards an understanding of the tertiary conformation of the protein. Since it is likely that such positions will be spatially clustered in the protein fold, these positions may give rise to useful distance constraints for substantiating model protein structures. The method is exemplified using data for a set of human mu class glutathione S-transferases. A novel aspect for predicting the behaviour of new polymorphic sequences is also discussed.

Reticulate evolution in molecular sequences is caused by recombination or gene conversion, and may interfere with the reconstruction of evolutionary history. This paper presents a program which calculates compatibility matrices for detecting reticulate evolution. In addition to visual inspection of matrices, they can be analysed statistically for clustering. The method is demonstrated using human and chimpanzee gamma-globin sequences.