Current Opportunities and Challenges of Next Generation Sequencing (NGS) of DNA; Determining Health and Diseases

Abstract

Many publications have demonstrated the huge potential of NGS methods in terms of new species \rdiscovery, environment monitoring, ecological studies, etc. [\r24,35,\r92,97,103\r]. Undoubtedly, NGS \rwill become one the major tools for species identification and for routine diagnostic use. While read \rlengths are still quite short for most existing systems ranging between 50 bp and 800 bp, they are \rlikely to improve soon. This will e\rnable easier, faster, and more reliable contig assembly and subsequent matching against reference databases. When data generation is no longer a \rbottleneck, the storage, speed of analysis, and interpretation of DNA sequence data are becoming \rthe major chal\rlenges. Also, the integration or the use of data originating from diverse datasets and \ra variety of data providers are serious issues that need to be addressed. Poor sequence record \rannotations and species name assignments are known problems that should be\rinstantly \raddressed and would allow the creation of reference databases used for routine diagnostics based \ron NGS. Samples with huge amounts of short DNA fragments need to be analyzed and compared \ragainst reference databases in an efficient and fast way. \rAlthough a number of solutions have \rbeen proposed by Industry; offering commercial software, there still remain hurdles to take. One of \rthe challenges that we need to address is data upload from client’s computers to central or \rdistributed data storage an\rd analysis services. Another one is the efficient parallelization of \ranalyses using cloud or grid solutions. The reliability and up\r-\rtime of storage and analyses facilities \ris another important problem that need to be addressed if one wants to use it for ro\rutine \rdiagnostics. Finally, the management, reporting and visualization of the analyses results are among \rthe last issues, but not the least challenging ones. Considering the constant growth of \rcomputational power and storage capacity needed by different b\rioinformatics applications, working \rwith single or a limited number of servers is no longer realistic. Using a cloud environment and grid \rcomputing is becoming a must. Even single cloud service provider can be restrictive for \rbioinformatics applications an\rd working with more than one cloud can make the workflow more \rrobust in the face of failures and always growing capacity needs. \rIn this white paper we review the \rcurrent state of the art in this field. We discuss the main limitations and challenges \rthat we\rneed to \raddress such as; data upload from client’s computers to central or distributed data storage and \ranalysis services; efficient parallelization of analyses using grid solutions; reliability and up\r-\rtime of \rstorage and analyses facilities for routine d\riagnostics; \rmanagement, retrieving and visualization of \rthe analyses results.