Next-generation sequencing in aging research: emerging applications, problems, pitfalls and possible solutions

Abstract

Recent technological advances that allow faster and cheaper DNA sequencing are now driving biological and medical research. In this review, we provide an overview of state-of-the-art next-generation sequencing (NGS) platforms and their applications, including in genome sequencing and resequencing, transcriptional profiling (RNA-Seq) and high-throughput survey of DNA-protein interactions (ChIP-Seq) and of the epigenome. Particularly, we focus on how new methods made possible by NGS can help unravel the biological and genetic mechanisms of aging, longevity and age-related diseases. In the same way, however, NGS platforms open discovery not available before, they also give rise to new challenges, in particular in processing, analyzing and interpreting the data. Bioinformatics and software issues plus statistical difficulties in genome-wide studies are discussed, as well as the use of targeted sequencing to decrease costs and facilitate statistical analyses. Lastly, we discuss a number of methods to gather biological insights from massive amounts of data, such as functional enrichment, transcriptional regulation and network analyses. Although in the fast-moving field of NGS new platforms will soon take center stage, the approaches made possible by NGS will be at the basis of molecular biology, genetics and systems biology for years to come, making them instrumental for research on aging.

Figure 1

Diagrammatic scheme of a NGS platform pipeline. Typically, DNA or RNA is fragmented into smaller pieces (A). Libraries are constructed from the fragments (or miRNAs) and sequenced at a high coverage (B). The sequenced reads are aligned to a reference genome (C) and the results are analyzed statistically and interpreted (D). Depending on the specific application, reads may be counted across genes (D1), SNPs detected (D2) or other analyses carried out.

Figure 2

NGS as a tool to uncover the genetic variation within (A) and across species (B). Within the humans species there is great variability in longevity and in susceptibility to age-related diseases, which to a large degree has a genetic basis. The capacity to resequence the genome of multiple individuals affected by a given disease or with different longevities offers powerful opportunities to identify polymorphisms and mutations that contribute to longevity and to age-related diseases (A). A large variance in lifespans is observed among similar species (B). For example, among primates, humans can live over 100 years, gorillas over 50 and Old World monkeys up to 40 (de Magalhaes, 2009). With the decreasing costs of sequencing, it is now possible to explore the genetic basis of such differences and identify coding or regulatory sequences that contribute to a given organism having a shorter or longer lifespan (B). Animals were drawn using fonts by Alan Carr.

Figure 3

Employing NGS platforms to study age-related changes. During the course of an organism’s lifetime, a number of genomic changes occur. NGS allows these changes to be quantified at a whole-genome level. Changes to be DNA, from single nucleotide mutations to large chromosome rearrangements, can be detected (A). Likewise, genome-wide epigenetic changes across the lifespan (or between different lifestyles or diets) can be assayed. Lastly, transcriptional changes with age can be quantified with unprecedented accuracy using NGS (C). Mouse and human figures were drawn using fonts by Alan Carr.

Figure 4

Moving from gene lists to biological insights. A number of data analysis methods are available to decrease the complexity of high-throughput data and gather biological information. In certain applications, transcriptional regulation information, either by incorporating existing knowledge or constructing transcriptional networks de novo, can be used to infer the causal structure of the data and infer regulatory nodes (A). Data can also be analyzed to identify new functional associations (B). For example, gene expression data can be used to cluster genes with similar profiles and identify putative functional interactions. Gene annotation information can also be integrated with experimental data to identify enriched functions and processes in large datasets (C). Lastly, data can be overlaid on protein-protein interaction networks to identify highly-connected hubs which tend to exert a regulating influence (D).