Incorporating Machine Learning into Established Bioinformatics Frameworks

The exponential growth of biomedical data in recent years has urged the application of numerous machine learning techniques to address emerging problems in biology and clinical research. By enabling the automatic feature extraction, selection, and generation of predictive models, these methods can be used to efficiently study complex biological systems. Machine learning techniques are frequently integrated with bioinformatic methods, as well as curated databases and biological networks, to enhance training and validation, identify the best interpretable features, and enable feature and model investigation. Here, we review recently developed methods that incorporate machine learning within the same framework with techniques from molecular evolution, protein structure analysis, systems biology, and disease genomics. We outline the challenges posed for machine learning, and, in particular, deep learning in biomedicine, and suggest unique opportunities for machine learning techniques integrated with established bioinformatics approaches to overcome some of these challenges.     

Simultaneous reconstruction of evolutionary history and epidemiological dynamics from viral sequences with the birth–death SIR model

The evolution of RNA viruses, such as human immunodeficiency virus (HIV), hepatitis C virus and influenza virus, occurs so rapidly that the viruses'' genomes contain information on past ecological dynamics. Hence, we develop a phylodynamic method that enables the joint estimation of epidemiological parameters and phylogenetic history. Based on a compartmental susceptible–infected–removed (SIR) model, this method provides separate information on incidence and prevalence of infections. Detailed information on the interaction of host population dynamics and evolutionary history can inform decisions on how to contain or entirely avoid disease outbreaks. We apply our birth–death SIR method to two viral datasets. First, five HIV type 1 clusters sampled in the UK between 1999 and 2003 are analysed. The estimated basic reproduction ratios range from 1.9 to 3.2 among the clusters. All clusters show a decline in the growth rate of the local epidemic in the middle or end of the 1990s. The analysis of a hepatitis C virus genotype 2c dataset shows that the local epidemic in the Córdoban city Cruz del Eje originated around 1906 (median), coinciding with an immigration wave from Europe to central Argentina that dates from 1880 to 1920. The estimated time of epidemic peak is around 1970.     

Identification,Phylogeny, and Comparative Expression of the Lipoxygenase Gene Family of the Aquatic Duckweed,Spirodela polyrhiza,during Growth and in Response to Methyl Jasmonate and Salt

Lipoxygenases (LOXs) (EC 1.13.11.12) catalyze the oxygenation of fatty acids and produce oxylipins, including the plant hormone jasmonic acid (JA) and its methyl ester, methyl jasmonate (MeJA). Little information is available about the LOX gene family in aquatic plants. We identified a novel LOX gene family comprising nine LOX genes in the aquatic plant Spirodela polyrhiza (greater duckweed). The reduced anatomy of S. polyrhiza did not lead to a reduction in LOX family genes. The 13-LOX subfamily, with seven genes, predominates, while the 9-LOX subfamily is reduced to two genes, an opposite trend from known LOX families of other plant species. As the 13-LOX subfamily is associated with the synthesis of JA/MeJA, its predominance in the Spirodela genome raises the possibility of a higher requirement for the hormone in the aquatic plant. JA-/MeJA-based feedback regulation during culture aging as well as the induction of LOX gene family members within 6 h of salt exposure are demonstrated.     

DNA Barcodes of Asian Houbara Bustard (Chlamydotis undulata macqueenii)

Populations of Houbara Bustards have dramatically declined in recent years. Captive breeding and reintroduction programs have had limited success in reviving population numbers and thus new technological solutions involving molecular methods are essential for the long term survival of this species. In this study, we sequenced the 694 bp segment of COI gene of the four specimens of Asian Houbara Bustard (Chlamydotis undulata macqueenii). We also compared these sequences with earlier published barcodes of 11 individuals comprising different families of the orders Gruiformes, Ciconiiformes, Podicipediformes and Crocodylia (out group). The pair-wise sequence comparison showed a total of 254 variable sites across all the 15 sequences from different taxa. Three of the four specimens of Houbara Bustard had an identical sequence of COI gene and one individual showed a single nucleotide difference (G > A transition at position 83). Within the bustard family (Otididae), comparison among the three species (Asian Houbara Bustard, Great Bustard (Otis tarda) and the Little Bustard (Tetrax tetrax)), representing three different genera, showed 116 variable sites. For another family (Rallidae), the intra-family variable sites among the individuals of four different genera were found to be 146. The COI genetic distances among the 15 individuals varied from 0.000 to 0.431. Phylogenetic analysis using 619 bp nucleotide segment of COI clearly discriminated all the species representing different genera, families and orders. All the four specimens of Houbara Bustard formed a single clade and are clearly separated from other two individuals of the same family (Otis tarda and Tetrax tetrax). The nucleotide sequence of partial segment of COI gene effectively discriminated the closely related species. This is the first study reporting the barcodes of Houbara Bustard and would be helpful in future molecular studies, particularly for the conservation of this threatened bird in Saudi Arabia.     

How much calcium to shell out? Eggshell calcium carbonate content is greater in birds with thinner shells,larger clutches and longer lifespans

The avian eggshell is a bio-ceramic structure that protects the embryo. It is composed almost entirely of calcium carbonate and a small amount of organic material. An optimal amount of calcium carbonate in the eggshell is essential for the embryo''s development, yet how the ratio of calcium carbonate to organic matter varies between species has not been investigated. Calcium is a limiting resource for most birds, so its investment in their eggs should be optimized for a bird''s life history. We measured the relative calcium carbonate content of eggshells in 222 bird species and tested hypotheses for how this trait has evolved with the life-history strategies of these species and other traits of their respective egg physiologies. We found that (i) eggshell calcium carbonate content was positively correlated with species having thinner eggshells and smaller than expected eggs relative to incubating parental mass, (ii) species with small mean clutch sizes had lower calcium carbonate content in their eggshells, and (iii) for species with larger clutch sizes, eggshell calcium carbonate content was negatively correlated with their mean lifespan. The pattern of lower eggshell calcium carbonate in longer lived, larger clutched birds suggests that calcium provision to the eggshell has long-term costs for the individual.     

Phylogenetics Applied to Genotype/Phenotype Association and Selection Analyses with Sequence Data from Angptl4 in Humans

Genotype/phenotype association analyses (Treescan) with plasma lipid levels and functional site prediction methods (TreeSAAP and PolyPhen) were performed using sequence data for ANGPTL4 from 3,551 patients in the Dallas Heart Study. Biological assays of rare variants in phenotypic tails and results from a Treescan analysis were used as “known” variants to assess the site prediction abilities of PolyPhen and TreeSAAP. The E40K variant in European Americans and the R278Q variant in African Americans were significantly associated with multiple lipid phenotypes. Combining TreeSAAP and PolyPhen performed well to predict “known” functional variants while reducing noise from false positives.     

Phylogenetic Relationships of Fungal Cytochromes c

The CYC1 gene encoding cytochrome c in the yeast Candida albicans was cloned by complementation of a cytochrome c-deficient mutant of Saccharomyces cerevisiae, and its DNA sequence was determined. The analysis of the amino acid sequences of cytochrome c from 14 fungal species and two isoforms from S. cerevisiae revealed sequences unique to fungi, and revealed a phylogenetic relationship with a pronounced divergence between Schizosaccharomyces pombe and other ascomycetous budding yeast. The C. albicans CYC1 cytochrome c sequence has been assigned Accession Number U57896 in the GenBank/EMBL database. © 1997 John Wiley & Sons, Ltd.     

The complete mitochondrial genome of Glycyphagus domesticus (Acari: Glycyphagidae) using next-generation sequencing: Insight into phylogeny of Acariformes

Glycyphagus domesticus (Acari: Glycyphagidae) is a pest mite with worldwide distribution. It damages stored products and produces allergens. Molecular diagnosis is important for the effective monitoring and control of stored-products and dust pest mites. The complete mitochondrial genome (mt genome) has become one of the most useful markers for molecular systematics, phylogenetics and diagnostics of stored-pest mites. In this study, next-generation sequencing was used to sequence the complete mt genome of G. domesticus. The complete mt genome was found to be 14,333 bp in size, containing 13 protein coding genes, 2 ribosomal RNA (rRNA) genes and 22 transfer RNA (tRNA) genes. We constructed phylogenetic trees based on concatenated thirteen protein-coding genes and two rRNAs genes. Our overall results showed that G. domesticus was found in its own separate clade from the 29 other mite species, which is likely due to the fact that this represents the first assembled mt genome of a member of the family Glycyphagidae. Overall, our study enriches the growing mt genome database for the Acariformes and will promote further studies.