Background

Revealing the origins of pathogenic viruses, crucial for cutting them off from the root and preventing future spillover, requires long-term hard work from scientists all around the world [1]. Although some infectious pathogens can be traced back decades, the debate on their origin continues. For example, AIDS was officially reported on June 5, 1981, by the Centers for Disease Control and Prevention of the USA. Five years later, HIV infection was detected in a human serum sample collected in Léopoldville in early 1959 [2]. Bayesian phylodynamic analyses using recovered viral gene sequences from decades-old paraffin-embedded tissues traced the most recent common ancestor (MRCA) of the M group of HIV back to approximately 1908 (CI 1884–1924), suggesting that HIV has been circulating in the human population for approximately 100 years [3]. MERS-CoV is another example, as it was first reported in a Saudi Arabian man in 2012 [4]. Bats are thought to be the reservoir hosts of MERS-CoV, and dromedary camels are considered to be the major intermediate host [5]; however, the transmission route from animals to humans is not well understood. Researchers tested 189 camel serum samples from 1983 to 1997 and found that 81% had neutralizing antibodies against MERS-CoV, suggesting long-term virus circulation in these animals [6]. Similarly, COVID-19 was first reported on December 27, 2019, in Wuhan, China [7,8], and the Huanan seafood market was suspected to be the place of origin [9]; however, disputes remain. Pekar and colleagues explored the evolutionary dynamics of the first wave of SARS-CoV-2 infections in China using a strict clock Bayesian phylodynamic analysis but failed to capture the index case [10], probably because the redundant sequences were not removed, which usually influences the accuracy of time of MRCA (tMRCA) estimation, as indicated in two recent tMRCA analysis [11,12]. Genome classification plays a critical role in tracing the origin of pathogenic viruses [3,12]. We have previously classified SARS-CoV-2 genomes based on two amino acids, Spike-614 and Orf8-84, and revealed 16 haplotypes. From those, three major haplotypes were found to separately drive the development of the pandemic in China and the world. However, genome classification based on amino acid mutations did not rule out recombination and reverse mutations. In this paper, we provide detailed protocols to filter and classify the massive number of viral genomes according to six highly linked mutations that happened in the early phase of the epidemic by custom scripts and common programs and estimate tMRCA of non-redundant genome subpools.