Viral recombination can generate novel genotypes with unique phenotypic characteristics, including transmissibility and virulence. Although the capacity for recombination among betacoronaviruses is well documented, recombination between strains of SARS-CoV-2 has not been characterized in detail. Here, we present a lightweight approach for detecting genomes that are potentially recombinant. This approach relies on identifying the mutations that primarily determine SARS-CoV-2 clade structure and then screening genomes for ones that contain multiple mutational markers from distinct clades. Among the over 537,000 genomes queried that were deposited on GISAID.org prior to February 16, 2021, we detected 1175 potential recombinant sequences. Using a highly conservative criteria to exclude sequences that may have originated through de novo mutation, we find that at least 30% (n = 358) are likely of recombinant origin. An analysis of deep-sequencing data for these putative recombinants, where available, indicated that the majority are high quality. Additional phylogenetic analysis and the observed co-circulation of predicted parent clades in the geographic regions of exposure further support the feasibility of recombination in this subset of potential recombinants. An analysis of these genomes did not reveal evidence for recombination hotspots in the SARS-CoV-2 genome. While most of the putative recombinant sequences we detected were genetic singletons, a small number of genetically identical or highly similar recombinant sequences were identified in the same geographic region, indicative of locally circulating lineages. Recombinant genomes were also found to have originated from parental lineages with substitutions of concern, including D614G, N501Y, E484K, and L452R. Adjusting for an unequal probability of detecting recombinants derived from different parent clades and for geographic variation in clade abundance, we estimate that at most 0.2-2.5% of circulating viruses in the US and UK are recombinant. Our identification of a small number of putative recombinants within the first year of SARS-CoV-2 circulation underscores the need to sustain efforts to monitor the emergence of new genotypes generated through recombination.

中文翻译：

---

大流行第一年重组 SARS-CoV-2 基因组的流通水平较低

病毒重组可以产生具有独特表型特征的新基因型，包括传播性和毒力。尽管 β 冠状病毒之间的重组能力已有充分记录，但 SARS-CoV-2 毒株之间的重组尚未得到详细描述。在这里，我们提出了一种用于检测可能重组的基因组的轻量级方法。这种方法依赖于识别主要决定 SARS-CoV-2 进化枝结构的突变，然后筛选基因组中包含来自不同进化枝的多个突变标记的突变。在 2021 年 2 月 16 日之前存放在 GISAID.org 上的超过 537,000 个基因组中，我们检测到了 1175 个潜在的重组序列。使用高度保守的标准排除可能源自从头突变的序列，我们发现至少 30% (n = 358) 可能源自重组。对这些假定重组体的深度测序数据（如果有）的分析表明，大多数都是高质量的。额外的系统发育分析和在暴露的地理区域中观察到的预测亲本进化枝的共循环进一步支持了在该潜在重组体子集中重组的可行性。对这些基因组的分析没有揭示 SARS-CoV-2 基因组中重组热点的证据。虽然我们检测到的大多数假定重组序列都是遗传单例，但在同一地理区域中鉴定出了少量遗传相同或高度相似的重组序列，表明存在本地流行的谱系。还发现重组基因组源自具有令人关注的取代的亲本谱系，包括 D614G、N501Y、E484K 和 L452R。调整检测来自不同亲本进化枝的重组体的不平等概率以及进化枝丰度的地理差异后，我们估计美国和英国最多 0.2-2.5% 的流行病毒是重组病毒。我们在 SARS-CoV-2 传播的第一年内鉴定出了少量假定的重组体，这强调需要持续努力监测通过重组产生的新基因型的出现。