Low-power portable devices for metagenomics analysis: Fog computing makes bioinformatics ready for the Internet of Things

On low‐power SoCs as storage bricks for Bioinformatics

“…As dataset for our test, we considered a set of 290 files, with a global size of 100 MB, derived from the sequencing of a metagenomic experiment . They are processed by DeepNano, which produces the same amount of files, with a global size of 4.6 MB in the FASTA format, to be processed by Kraken to finally obtain 790 tuples with a global size of 3.2 MB …”

“…In the case‐study considered in this paper, SoCs devices are used for a real‐time analysis of the information streamed by portable sequencing machines, as to monitor bacterial communities and provide feedback and/or alarms in case of deviations. This solution, proposed in our previous work, envisages a possibly large amount of data provided by genome sequencing to be managed via SoCs. Hence, the interest in exploring and assessing the data and metadata performances of a file system in support of such application, which is the scope of the present work.…”

“…In a previous work by Merelli et al, the authors demonstrated the feasibility and advantages of executing such a metagenomic analysis in SoCs within an edge‐fog computing scenario, which is true also for some compute‐intensive applications, such as the N‐Body algorithm, when considering performance, energy consumption, and economical aspects altogether …”

“…Among the various real‐life use‐cases analysed in the context of COSA, a success story is represented by the blueprint of a simple and portable approach for genome sequencing, which exploits low‐power SoCs in a so‐called Fog computing paradigm for processing locally the raw data so to send only meaningful results over Internet . Such scenario is of practical relevance in a growing number of situations, given the use of portable sequencing machines such as the Oxford Nanopore MinION has become extremely common, eg, checking specific bacteria in air‐filters or waters and triggering alarms in case specific and dangerous bacteria are found, monitoring pollution signals to identifying and maintaining a set point of the bacterial communities in soil (eg, for cultivated lands, greenhouses, and animal farms), providing rapid feedback in case of deviations.…”

New High-Tech Flexible Networks for the Monitoring of Deep-Sea Ecosystems

“…Similarly, ref 103 used a random forest based approach for selecting the relevant biomarkers for classification of ocean, harbor, and ballast water samples. 104 used a deep recurrent neural network (approach for a base calling application on portable sequencing machines, 105 where meaningful results were sent to a cloud service through an Internet of Things framework for further analysis. 106 Along with molecular-based monitoring tools, other chemical sensing applications may complement DNA probing 107 and sequencing.…”