Analysis of high-dimensional genomic data using MapReduce based probabilistic neural network

BACKGROUND

The size of genomics data has been growing rapidly over the last decade. However, the conventional data analysis techniques are incapable of processing this huge amount of data. For the efficient processing of high dimensional datasets, it is essential to develop some new parallel methods.

METHODS

In this work, a novel distributed method is presented using Map-Reduce (MR)-based approach. The proposed algorithm consists of MR-based Fisher score (mrFScore), MR-based ReliefF (mrRefiefF), and MR-based probabilistic neural network (mrPNN) using a weighted chaotic grey wolf optimization technique (WCGWO). Here, mrFScore, and mrRefiefF methods are introduced for feature selection (FS), and mrPNN is implemented as an effective method for microarray classification. The proper choice of smoothing parameter (σ) plays a major role in the prediction ability of the PNN which is addressed using a novel technique namely, WCGWO. The WCGWO algorithm is used to select the optimal value of σ in PNN.

RESULTS

These algorithms have been successfully implemented using the Hadoop framework. The proposed model is tested by using three large and one small microarray datasets, and a comparative analysis is carried out with the existing FS and classification techniques. The results suggest that WCGWO-mrPNN can outperform other methods for high dimensional microarray classification.

CONCLUSION

The effectiveness of the proposed methods are compared with other existing schemes. Experimental results reveal that the proposed scheme is accurate and robust. Hence, the suggested scheme is considered to be a reliable framework for microarray data analysis.

SIGNIFICANCE

Such a method promotes the application of parallel programming using Hadoop cluster for the analysis of large-scale genomics data, particularly when the dataset is of high dimension.

Chaotic emperor penguin optimised extreme learning machine for microarray cancer classification

Microarray technology plays a significant role in cancer classification, where a large number of genes and samples are simultaneously analysed. For the efficient analysis of the microarray data, there is a great demand for the development of intelligent techniques. In this article, the authors propose a novel hybrid technique employing Fisher criterion, ReliefF, and extreme learning machine (ELM) based on the principle of chaotic emperor penguin optimisation algorithm (CEPO). EPO is a recently developed metaheuristic method. In the proposed method, initially, Fisher score and ReliefF are independently used as filters for relevant gene selection. Further, a novel population-based metaheuristic, namely, CEPO was proposed to pre-train the ELM by selecting the optimal input weights and hidden biases. The authors have successfully conducted experiments on seven well-known data sets. To evaluate the effectiveness, the proposed method is compared with original EPO, genetic algorithm, and particle swarm optimisation-based ELM along with other state-of-the-art techniques. The experimental results show that the proposed framework achieves better accuracy as compared to the state-of-the-art schemes. The efficacy of the proposed method is demonstrated in terms of accuracy, sensitivity, specificity, and F-measure.

Characterization of oceanic Noctiluca blooms not associated with hypoxia in the Northeastern Arabian Sea

Intense blooms of the heterotrophic dinoflagellate, green Noctiluca scintillans, have been reported annually in the Northern Arabian Sea since the early 2000s. Although not known to produce organic toxins, these blooms are still categorized as a harmful due to their association with massive fish mortalities. Recent work has attributed these blooms to the vertical expansion of the oxygen minimum zone, driven by cultural eutrophication from major coastal cities in western India. As diatoms are preferred prey of green Noctiluca scintillans, more frequent blooms of this mixotroph will likely impact the productivity of important fisheries in the region. The present study uses a satellite algorithm to determine the distribution of both diatom and green Noctiluca blooms in the Northeastern Arabian Sea from 2009 to 2016. The results from shipboard microscopy of phytoplankton community composition were used to validate the satellite estimates. The satellite algorithm showed 76% accuracy for detection of green Noctiluca and 92% for diatoms. Shipboard measurements and data from biogeochemical-Argo floats were used to assess the relationship between oxygen concentrations and green Noctiluca blooms in the Northeastern Arabian Sea. Regardless of the presence of a Noctiluca bloom, the dissolved oxygen in the photic zone was always >70% saturated, with an average oxygen saturation >90%. The variability in the relative abundance of diatoms and green Noctiluca is not correlated with changes in oxygen concentration. These findings provide no evidence that cultural eutrophication has contributed to the decadal scale shifts in plankton composition in the Northeastern Arabian Sea oceanic waters. Conversely, the climatic warming of surface waters would have intensified stratification, thereby reducing net nutrient flux to the photic zone and decreasing silicate to nitrate ratios (Si:N); both factors that could increase the competitive advantage of the mixotroph, green Noctiluca, over diatoms. If so, the decadal-scale trajectory of phytoplankton community composition in the Northeastern Arabian Sea may be a harbinger of future climate-driven change in other productive oceanic systems.

Response of phytoplankton community and size classes to green Noctiluca bloom in the northern Arabian Sea

A comprehensive analysis on the phytoplankton ecology with special reference to different phytoplankton size classes was carried out at green Noctiluca scintillans (hereafter Noctiluca) bloom and non-bloom locations in offshore waters of the northern Arabian Sea. At the bloom locations, green Noctiluca represented a dense mono-specific proliferation with average cell density of 10.16 ± 5.806 × 10⁴ cells-L^-1 and relative abundance share of 98.63%. Active photosynthesis through prasinophytic endosymbiont was depicted from net community production magnitude reaching 85.26 mgC/m³/Day under low prey abundance. Parallel swarming of Porpita porpita, a voracious copepod feeder signified the competitive advantage of Noctiluca to have the phytoplankton prey. Average concentration of picophytoplankton biomass was eleven times lower in surface waters of non-bloom stations in comparison to bloom. Higher N:P ratio in subsurface waters of non-bloom stations signified non-utilization of nitrogenous nutrients. Green Noctiluca bloom onset subsequent to diatom rich conditions was evident from spatio-temporal ocean colour satellite imageries.

An optical remote sensing approach for ecological monitoring of red and green Noctiluca scintillans

An ecosystem disruptive bloom of red Noctiluca scintillans (hereafter Noctiluca) was observed in coastal waters of the north-western Bay of Bengal during April 2014. Based on the principle of phytoplankton group/species specific remote sensing reflectance (R_rs), a technique of detecting green Noctiluca and diatom was developed earlier using R_rs at 443, 488, and 531 nm of Moderate Imaging Spectroradiometer-Aqua (MODIS). This was appropriately modified to detect bloom of red Noctiluca in coastal waters of the Bay of Bengal. Additional R_rs data at longer wavelengths viz. 667 and 678 nm were included in the existing algorithm, and the spectral shapes were accounted to detect the bloom of red Noctiluca. The classification scheme discriminates red Noctiluca from the green form of the same species and diatom. Phytoplankton group/species products were generated using the modified approach and validated with the reported events of red and green Noctiluca blooms in the Indian coastal waters. The present study also highlights two specific results based on MODIS retrieved time-series phytoplankton group/species image analysis: first, the observation of coexistence of diatom, red, and green Noctiluca in coastal waters of the north-western Bay of Bengal, and the second, phytoplankton community shift resulting in red/green Noctiluca proliferation following diatom.

Effect of physico-chemical regimes and tropical cyclones on seasonal distribution of chlorophyll-a in the Chilika Lagoon, east coast of India

A comprehensive analysis on spatiotemporal variation in physico-chemical variables and their control on chlorophyll-a during 2013-14 was carried out in the Chilika Lagoon. Spatiotemporal variation in physico-chemical regimes significantly controlled the phytoplankton biomass of the lagoon. Further, precipitation-induced river/terrestrial freshwater influx and marine influence controlled the physico-chemical regimes of the Chilika Lagoon, such as nutrients (NH₄⁺, NO₃^-, NO₂^-, PO₄^3- and Si(OH)₄), temperature, salinity, total suspended matter and dissolved oxygen. This study revealed significant effects of tropical cyclones Phailin (2013) and Hudhud (2014) on physico-chemical regimes and in turn the phytoplankton biomass of the lagoon. Although both cyclones Phailin (2013) and Hudhud (2014) were intense, Phailin (2013) had a greater impact on the Chilika Lagoon due to the proximity of its landfall. Heavy precipitation caused an influx of nutrient-rich freshwater, both during each cyclone's passage, through rainfall, and after, through increased river flow and terrestrial run-off. The increase in nutrients, carried by the run-off, promoted phytoplankton growth, albeit in lag phase. In general, phytoplankton growth was controlled by nitrogenous nutrients. However, the addition of SiO₄ through terrigenous run-off fuelled preferential growth of diatoms. The salinity pattern (which can be considered a proxy for fresh and marine water influx) indicated injection of freshwater nutrients into the northern, southern and central sectors of the lagoon through riverine/terrestrial freshwater run-off; marine influx was restricted to the mouth of the lagoon. Present and past magnitudes of salinity and chlorophyll-a were also compared to better understand the pattern of variability. A significant change in salinity pattern was noticed after the opening of an artificial inlet, because of the resulting higher influx of marine water. The overall phytoplankton biomass (using chlorophyll-a concentration as a proxy) remained consistent in the lagoon pre- and post-restoration. Due to the wide range of salinity and temperature tolerance, diatoms remained dominant in both pre- and post-restoration periods, but the overall phytoplankton diversity increased after the artificial inlet was dredged.

Environmental dynamics of red Noctiluca scintillans bloom in tropical coastal waters

An intense bloom of red Noctiluca scintillans (NS) occurred off the Rushikulya estuarine region along the east coast of India, an important site for mass nesting events of the vulnerable Olive Ridley sea turtle. At its peak, densities of NS were 3.3×10(5) cells-l(-1), with low relative abundance of other phytoplankton. The peak bloom coincided with high abundance of gelatinous planktivores which may have facilitated bloom development by their grazing on other zooplankton, particularly copepods. Ammonium concentrations increased by approximately 4-fold in the later stages of bloom, coincident with stable NS abundance and chlorophyll concentrations in the nano- and microplankton. This increase likely was attributable to release of intracellular ammonium accumulated through NS grazing. Dissolved oxygen concentrations decreased in sub-surface waters to near hypoxia. Micro-phytoplankton increasingly dominated chlorophyll-a biomass as the bloom declined, with diminishing picoplankton abundance likely the result of high predation by the ciliate Mesodinium rubrum. Together, these data illustrate factors that can disrupt ecosystem balance in this critically important Indian coastal region.

Distribution of trace metals in surface seawater and zooplankton of the Bay of Bengal, off Rushikulya estuary, East Coast of India

Concentrations of trace metals such as iron (Fe), copper (Cu), zinc (Zn), cobalt (Co), nickel (Ni), manganese (Mn), lead (Pb), cadmium (Cd), chromium (Cr), arsenic (As), vanadium (V), and selenium (Se) were determined in seawater and zooplankton from the surface waters off Rushikulya estuary, north-western Bay of Bengal. During the study period, the concentration of trace metals in seawater and zooplankton showed significant spatio-temporal variation. Cu and Co levels in seawater mostly remained non-detectable. Other elements were found at higher concentrations and exhibited marked variations. The rank order distribution of trace metals in terms of their average concentration in seawater was observed as Fe>Ni>Mn>Pb>As>Zn>Cr>V>Se>Cd while in zooplankton it was Fe>Mn>Cd>As>Pb>Ni>Cr>Zn>V>Se. The bioaccumulation factor (BAF) of Fe was highest followed by Zn and the lowest value was observed with Ni. Results of correlation analysis discerned positive affinity and good relationship among the majority of the trace metals, both in seawater and zooplankton suggesting their strong affinity and coexistence.

Phytoplankton community structure in local water types at a coastal site in north-western Bay of Bengal

A comprehensive analysis on seasonal distribution of phytoplankton community structure and their interaction with environmental variables was carried out in two local water types (type 1 < 30 m isobath and Type 2 > 30 m isobath) at a coastal site in north-western Bay of Bengal. Phytoplankton community was represented by 211 taxa (146 marine, 37 fresh, 2 brackish, 20 marine-fresh, and 6 marine-brackish-fresh) belonging to seven major groups including 45 potential bloom forming and 22 potential toxin producing species. The seasonal variability depicted enrichment of phytoplankton during pre-monsoon in both water types. Total phytoplankton abundance pattern observed with inter-annual shift during monsoon and post-monsoon period at both water types. In both water types, diatom predominance was observed in terms of species richness and abundance comprising of centric (82 sp.) and pennate (58 sp.) forms. Pennate diatoms, Thalassiothrix longissima and Skeletonema costatum preponderated in both the water types. The diatom abundance was higher in type 1 in comparison to type 2. In general, SiO4 found to fuel growth of the dominant phytoplankton group, diatom in both the water types despite comparative lower concentration of other macronutrients in type 2.

Spatio-temporal distribution of chlorophyll-a in relation to physico-chemical parameters in coastal waters of the northwestern Bay of Bengal

The present study focuses on understanding the long-term distribution of physico-chemical parameters and their influence on the distribution of chlorophyll-a (chl-a) at a coastal site in the northwestern Bay of Bengal. Chl-a showed large variability (0.12 to 10.05 mg m(-3)) on a spatio-temporal scale during the study period. However, the distribution showed a similar pattern with marginal variability from March 2010 to February 2011 and March 2011 to February 2012. The vertical distribution of salinity, pH, total suspended matter (TSM) and chl-a showed systematic temporal variability. However, dissolved oxygen (DO) and nutrients (nitrite + nitrate, phosphate, silicate) did not show any significant spatio-temporal trend. Chl-a showed bimodal distribution on an annual scale, with the first peak appearing during the pre-monsoon period in March due to a seasonal phytoplankton bloom, whereas the second peak occurring during September as a result of nutrient loading from river influx due to monsoonal precipitation. Factor analysis revealed the association of low salinity and high nutrients with chl-a. This infers that the nutrients brought by the influx of river into the study area were fuelling the growth and abundance of phytoplankton. Cluster analysis resulted in two distinct clusters among all physico-chemical datasets, indicating the presence of two distinct areas separated by the 30 m isobath that were strongly influenced by physico-chemical characteristics associated with the seasonal monsoon.

Distribution of hydro-biological parameters in coastal waters off Rushikulya Estuary, East Coast of India: a premonsoon case study

The hydro-biological parameters of coastal waters off Rushikulya estuary was investigated during premonsoon 2011. Important hydro-biological parameters such as water temperature, salinity, pH, DO, NO2, NO3, NH4, PO4, SiO4, TSM, Chl-a, phytoplankton and zooplankton were measured during the present study. Temperature established a strong positive correlation with salinity and pH during the present study. Chl-a found in positive relation with NO3, SiO, and TSM. Analysis of variance revealed significant monthly variation in pH, salinity and TSM. Significant station wise variation was observed in DO and most of the nutrients i.e., NO3, NH4, PO4, SiO4. A total of 119 species of phytoplankton were identified of which 84 species are of diatoms, 22 species of dinoflagellates, 7 species of green algae, 5 species of cyanobacteria (blue green algae) and 1 species of cocolithophore. Phytoplankton abundance varied between 25543 (Nos. L(-1)) and 36309 (Nos. L(-1)). Diatoms dominated the phytoplankton community followed by dinoflagellates in all the months. Diatoms contributed to 82-89% of the total phytoplankton population density whereas dinoflagellates contributed to 6-12%. The regression between Chl-a and phytoplankton abundance resulted with weak relation (R(2) = 0.042). Zooplankton fauna composed of 134 species of holoplankton and 20 types of meroplankton were encountered during the study period. Zooplankton population dominated by copepod during all months and accounted for 74 to 85% to the total zooplankton. The population density ranged from 6959 to 35869 Nos./10 m(3). Analysis of variance explained no significant variation in total zooplankton abundance and also for different groups of zooplankton.