High-throughput or next-generation sequencing (NGS) technologies have become an established and affordable experimental framework in biological and medical sciences for all basic and translational research. Processing and analyzing NGS data is challenging. NGS data are big, heterogeneous, sparse, and error prone. Although a plethora of tools for NGS data analysis has emerged in the past decade, (i) software development is still lagging behind data generation capabilities, and (ii) there is a ‘cultural’ gap between the end user and the developer.
Generic software template libraries specifically developed for NGS can help in dealing with the former problem, whilst coupling template libraries with visual programming may help with the latter. Here we scrutinize the state-of-the-art low-level software libraries implemented specifically for NGS and graphical tools for NGS analytics. An ideal developing environment for NGS should be modular (with a native library interface), scalable in computational methods (i.e. serial, multithread, distributed), transparent (platform-independent), interoperable (with external software interface), and usable (via an intuitive graphical user interface). These characteristics should facilitate both the run of standardized NGS pipelines and the development of new workflows based on technological advancements or users’ needs. We discuss in detail the potential of a computational framework blending generic template programming and visual programming that addresses all of the current limitations.
In the long term, a proper, well-developed (although not necessarily unique) software framework will bridge the current gap between data generation and hypothesis testing. This will eventually facilitate the development of novel diagnostic tools embedded in routine healthcare.