Long-read sequencing technologies have now reached a level of accuracy and yield that allows their application to variant detection at a scale of tens to thousands of samples. Concomitant with the development of new computational tools, the first population-scale studies involving long-read sequencing have emerged over the past 2 years and, given the continuous advancement of the field, many more are likely to follow. In this Review, we survey recent developments in population-scale long-read sequencing, highlight potential challenges of a scaled-up approach and provide guidance regarding experimental design. We provide an overview of current long-read sequencing platforms, variant calling methodologies and approaches for de novo assemblies and reference-based mapping approaches. Furthermore, we summarize strategies for variant validation, genotyping and predicting functional impact and emphasize challenges remaining in achieving long-read sequencing at a population scale.

长读长测序技术现已达到一定的准确度和产量水平，可应用于数十至数千个样本规模的变异检测。随着新计算工具的发展，在过去两年中出现了第一个涉及长读长测序的群体规模研究，并且鉴于该领域的不断进步，可能会有更多的研究跟进。在这篇综述中，我们调查了群体规模长读长测序的最新进展，强调了扩大规模方法的潜在挑战，并提供了有关实验设计的指导。我们概述了当前的长读长测序平台、变体调用方法以及从头组装的方法和基于参考的映射方法。此外，我们总结了变异验证、基因分型和预测功能影响的策略，并强调了在人群规模上实现长读长测序所面临的挑战。