A fast hierarchical clustering algorithm for large-scale protein sequence data sets

Genomic analysis of Gordonia polyisoprenivorans strain R9, a highly effective 17 beta-estradiol- and steroid-degrading bacterium

The increasing levels of estrogens and pollution by other steroids pose considerable challenges to the environment. In this study, the genome of Gordonia polyisoprenivorans strain R9, one of the most effective 17 beta-estradiol- and steroid-degrading bacteria, was sequenced and annotated. The circular chromosome of G. polyisoprenivorans R9 was 6,033,879 bp in size, with an average GC content of 66.91%. More so, 5213 putative protein-coding sequences, 9 rRNA, 49 tRNA, and 3 sRNA genes were predicted. The core-pan gene evolutionary tree for the genus Gordonia showed that G. polyisoprenivorans R9 is clustered with G. polyisoprenivorans VH2 and G. polyisoprenivorans C, with 93.75% and 93.8% similarity to these two strains, respectively. Altogether, the three G. polyisoprenivorans strains contained 3890 core gene clusters. Strain R9 contained 785 specific gene clusters, while 501 and 474 specific gene clusters were identified in strains VH2 and C, respectively. Furthermore, whole genome analysis revealed the existence of the steroids and estrogens degradation pathway in the core genome of all three G. polyisoprenivorans strains, although the G. polyisoprenivorans R9 genome contained more specific estrogen and steroid degradation genes. In strain R9, 207 ABC transporters, 95 short-chain dehydrogenases (SDRs), 26 monooxygenases, 21 dioxygenases, 7 aromatic ring-hydroxylating dioxygenases, and 3 CoA esters were identified, and these are very important for estrogen and steroid transport, and degradation. The results of this study could enhance our understanding of the role of G. polyisoprenivorans R9 in estradiol and steroid degradation as well as evolution within the G. polyisoprenivorans species.

Genome-Enhanced Detection and Identification (GEDI) of plant pathogens

Plant diseases caused by fungi and Oomycetes represent worldwide threats to crops and forest ecosystems. Effective prevention and appropriate management of emerging diseases rely on rapid detection and identification of the causal pathogens. The increase in genomic resources makes it possible to generate novel genome-enhanced DNA detection assays that can exploit whole genomes to discover candidate genes for pathogen detection. A pipeline was developed to identify genome regions that discriminate taxa or groups of taxa and can be converted into PCR assays. The modular pipeline is comprised of four components: (1) selection and genome sequencing of phylogenetically related taxa, (2) identification of clusters of orthologous genes, (3) elimination of false positives by filtering, and (4) assay design. This pipeline was applied to some of the most important plant pathogens across three broad taxonomic groups: Phytophthoras (Stramenopiles, Oomycota), Dothideomycetes (Fungi, Ascomycota) and Pucciniales (Fungi, Basidiomycota). Comparison of 73 fungal and Oomycete genomes led the discovery of 5,939 gene clusters that were unique to the targeted taxa and an additional 535 that were common at higher taxonomic levels. Approximately 28% of the 299 tested were converted into qPCR assays that met our set of specificity criteria. This work demonstrates that a genome-wide approach can efficiently identify multiple taxon-specific genome regions that can be converted into highly specific PCR assays. The possibility to easily obtain multiple alternative regions to design highly specific qPCR assays should be of great help in tackling challenging cases for which higher taxon-resolution is needed.

Improved multi-objective clustering algorithm using particle swarm optimization

Multi-objective clustering has received widespread attention recently, as it can obtain more accurate and reasonable solution. In this paper, an improved multi-objective clustering framework using particle swarm optimization (IMCPSO) is proposed. Firstly, a novel particle representation for clustering problem is designed to help PSO search clustering solutions in continuous space. Secondly, the distribution of Pareto set is analyzed. The analysis results are applied to the leader selection strategy, and make algorithm avoid trapping in local optimum. Moreover, a clustering solution-improved method is proposed, which can increase the efficiency in searching clustering solution greatly. In the experiments, 28 datasets are used and nine state-of-the-art clustering algorithms are compared, the proposed method is superior to other approaches in the evaluation index ARI.

A modified two-stage Markov clustering algorithm for large and sparse networks

BACKGROUND

Graph-based hierarchical clustering algorithms become prohibitively costly in both execution time and storage space, as the number of nodes approaches the order of millions.

OBJECTIVE

A fast and highly memory efficient Markov clustering algorithm is proposed to perform the classification of huge sparse networks using an ordinary personal computer.

METHODS

Improvements compared to previous versions are achieved through adequately chosen data structures that facilitate the efficient handling of symmetric sparse matrices. Clustering is performed in two stages: the initial connected network is processed in a sparse matrix until it breaks into isolated, small, and relatively dense subgraphs, which are then processed separately until convergence is obtained. An intelligent stopping criterion is also proposed to quit further processing of a subgraph that tends toward completeness with equal edge weights. The main advantage of this algorithm is that the necessary number of iterations is separately decided for each graph node.

RESULTS

The proposed algorithm was tested using the SCOP95 and large synthetic protein sequence data sets. The validation process revealed that the proposed method can reduce 3-6 times the processing time of huge sequence networks compared to previous Markov clustering solutions, without losing anything from the partition quality.

CONCLUSIONS

A one-million-node and one-billion-edge protein sequence network defined by a BLAST similarity matrix can be processed with an upper-class personal computer in 100 minutes. Further improvement in speed is possible via parallel data processing, while the extension toward several million nodes needs intermediary data storage, for example on solid state drives.

Evaluation of BLAST-based edge-weighting metrics used for homology inference with the Markov Clustering algorithm

Background

Clustering protein sequences according to inferred homology is a fundamental step in the analysis of many large data sets. Since the publication of the Markov Clustering (MCL) algorithm in 2002, it has been the centerpiece of several popular applications. Each of these approaches generates an undirected graph that represents sequences as nodes connected to each other by edges weighted with a BLAST-based metric. MCL is then used to infer clusters of homologous proteins by analyzing these graphs. The various approaches differ only by how they weight the edges, yet there has been very little direct examination of the relative performance of alternative edge-weighting metrics. This study compares the performance of four BLAST-based edge-weighting metrics: the bit score, bit score ratio (BSR), bit score over anchored length (BAL), and negative common log of the expectation value (NLE). Performance is tested using the Extended CEGMA KOGs (ECK) database, which we introduce here.

Results

All metrics performed similarly when analyzing full-length sequences, but dramatic differences emerged as progressively larger fractions of the test sequences were split into fragments. The BSR and BAL successfully rescued subsets of clusters by strengthening certain types of alignments between fragmented sequences, but also shifted the largest correct scores down near the range of scores generated from spurious alignments. This penalty outweighed the benefits in most test cases, and was greatly exacerbated by increasing the MCL inflation parameter, making these metrics less robust than the bit score or the more popular NLE. Notably, the bit score performed as well or better than the other three metrics in all scenarios.

Conclusions

The results provide a strong case for use of the bit score, which appears to offer equivalent or superior performance to the more popular NLE. The insight that MCL-based clustering methods can be improved using a more tractable edge-weighting metric will greatly simplify future implementations. We demonstrate this with our own minimalist Python implementation: Porthos, which uses only standard libraries and can process a graph with 25 m + edges connecting the 60 k + KOG sequences in half a minute using less than half a gigabyte of memory.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-015-0625-x) contains supplementary material, which is available to authorized users.