SARS-CoV-2 nucleic acid positive samples were collected in Weihai from March 2022 to March 2023. These samples consisted of throat swab samples from local cases and foreign imported cases. Samples of local cases were provided by hospital clinical laboratories, district Centers for Disease Control Prevention and Control (CDC) and third-party testing laboratories.Samples of foreign imported cases were provided by Weihai Customs Laboratory. These samples were tested positive by using Real-time PCR before we received them. We retested these samples and selected positive samples with CT < 32 for whole genome sequencing.

The viral RNA was extracted by Automatic nucleic acid extraction instrument (Jiangsu Bioperfectus Technologies Co,Ltd). Using the viral RNA as a template, the whole genome of SARS-CoV-2 was amplified by PCR using the 2019-nCoV Genome Capture Kit (MicroFuture, Beijing, China). The PCR products were purified using AMPure XP magnetic beads. Library construction was performed using the Nextera XT DNA Library Preparation Kit and Nextera XT Index Kit (Illumina, San Diego, CA, United States). Whole genome sequencing was performed using Illumina Nextseq2000 sequencer and NextSeq 1000/2000 P1 reagents (300 cycles) (MS-102-2002).The sequencing read length was set as 2 × 151.

SARS-CoV-2 genome aligment and assembly were performed using CLC Genomics Workbench version22.0. Wuhan-Hu-1 (NC_045512.2) was used as the reference sequence, and QC analysis, trim and map were performed on the FASTQ file to obtain the spliced SARS-CoV-2 whole genome sequence. Nextclade (15) (https://clades.nextstrain.org/ Version2.13.0) was used to perform clade typing, pangolin typing, QC, and mutation site analysis on the obtained SARS-CoV-2 whole genome sequences. Based on the typing results, the relevant sequences were downloaded from the GISAID database, and the phylogenetic tree was constructed using MEGA version11.The Neighbor-Joining (NJ) method was used to construct the evolutionary tree and the Bootstrap value was set to 1,000 times to evaluate the reliability of the phylogenetic tree.

BF.7 and BA.5.2 sequences collected from Weihai after December 2022 was performed phylodynamic analysis.Firstly, PhyloSuite (16) version v1.2.212 was used to match the best substitution model. Based on the analysis results, the best substitution model was TIM with 4 Gamma category count.Then TempEst (17) version 1.5.313 was used to detect whether there was enough temporal molecular signal for phylodynamic analysis. The analysis results indicated that after removing some outliers, there was enough temporal molecular signal for subsequent analysis.BEATUti was used for parameter setting, set the clock model to strict molecular clock model, set the tree priors to coalescent Bayesian skyline, and set the chain length to 100million (18). We used BEAST version 2.7.4 to run the XML file generated by BEATUti.Tracer version 1.7.2 was used to check the convergence of MCMC chains (ESS > 200) and performed Bayesian Skyline Analysis to estimate the effective population size for both BA.5.2 and BF.7. TreeAnnotator (19) was used to remove 10% of the MCMC chains as burn-in to build a maximum clade credibility tree (MCC tree). FigTree version1.4.4 was used to add time axis to the MCC tree.

We collected SARS-CoV-2 nucleic acid positive samples from March 2022 to March 2023 in Weihai, and selected samples with CT < 32 for whole-genome sequencing. We obtained a total of 704 SARS-CoV-2 whole-genome sequences. After removing some duplicate sequences and low-quality sequences, a total of 581 whole genome sequences were finally obtained for subsequent analysis. Among them, 553 sequences were local cases sequences and 28 were foreign imported cases sequences. The typing results showed that 568 sequences were Omicron variants and 13 sequences were Gamma variants. According to the Clade typing results, there were eight clades, including 22B (71.43%), 21 L (20.83%), 21 K (3.27%), 20 J (2.24%), 22C (0.86%), 22F (0.69%), 22D (0.52%), and 22A (0.17%). According to the Pangolin typing results, there were 30 Pango lineages, the predominant lineages were BF.7.14 (29.43%), BA.5.2.48 (25.82%), BA.2 (18.59%), and the other lineages accounting for more than 1% included BA.5.2.1 (6.20%), BA.1.1 (3.27%), BA.5.2 (2.93%), BA.7.14.1 (2.75%), P.1.15 (2.24%), and BA.5.2.49 (1.55%) (Figure 1A).

From March 2022 to March 2023, we obtained 28 SARS-CoV-2 whole genome sequences of foreign imported cases. According to the Clade typing results, there were 5 clades, including 22B, 22F, 22D, 21 L, 22A and 11 Pangolin lineages, including BA.5.2, BA.5.2.1, BA.5.2.20, BA.5.2.44, BA.5.1.15, XBB.1, XBB.1.4, BN.1.3, BA.2, BA.2.3, BA.4.1 (Figure 1B). BA.5.2 and its descendant lineages were dominant in the imported cases (64.28%).The Pangolin lineages of imported cases were generally consistent with those detected in local cases during the same period.However,some rare variants were also detected in the imported cases, such as BA.4.1, XBB.1, BA.5.1.15, BN.1.3, which did not appear in the local cases.

According to the temporal distribution of SARS-CoV-2 variants in Weihai, we could find that from March to April 2022, the dominant strains in Weihai were BA. 2. From May to November 2022, the dominant strains were BA. 1.1, P.1.15, and BA. 5.2.From December 2022 to March 2023, BA.5.2 and its descendant lineages, as well as BF.7 and its descendant lineages, became the dominant strains in Weihai (Figure 2).

From March 2022 to November 2022, before the adjustment of epidemic policies,we obtained a total of 277 SARS-CoV-2 whole genome sequences, including 251 local cases and 26 foreign imported cases. The variants with the highest proportion were BA.2 and its descendant lineages (45.49%). Due to the strict epidemic prevention policy, the COVID-19 cases in Weihai during this period were mainly local clusters. During this period, one large-scale local clusters and three small-scale clusters occurred in Weihai.These four clusters were caused by four different Pangolin lineages, including BA. 2, BA. 1.1, P.1.15 and BA. 5.2.1. We conducted phylogenetic analysis on the related sequences involved in these clusters, and the phylogenetic tree showed that:In the same cluster, almost all local cases sequences were on the same branch, and the gene diversities between these sequences were very limited, indicating that the sequences were highly homologous (Figure 3). The alignment results showed that their whole genome sequences were almost identical, and the SNP differences between them were no more than 3.In addition,By comparing with the imported cases in the same period.it can be found that the imported cases have the same pangolin lineages with the local cases.But they were not on the same branch in phylogenetic tree.This indicated that these four local clusters were not caused by these imported cases.By comparing with the homologous sequences downloaded from the GISAID database, we found that the local clusters of BA.2, BA.1.1 and BA.5.2.1 occurred at roughly the same time as the homologous sequences uploaded in the GISAID database. This indicated that the variants that triggered these three clusters are roughly the same as the SARS-CoV-2 variants that were circulating globally at the same time. The only exception was P.1.15, the local cluster of the P.1.15 variant occurred in June 2022, while most of the homologous GISAID database sequences were uploaded from March to July 2021.

After December 2022, China adjusted its epidemic prevention policies and new infection cases increased rapidly in the short term,so whole-genome sequencing for SARS-Cov-2 variants was changed to once a week in Weihai. From December 2022 to March 2023, we obtained a total of 304 SARS-CoV-2 whole genome sequences.Among them, BF.7 and BA.5.2 were the two dominant variants in Weihai.BF.7 and its descendant lineages, including BF.7.14, BF.7.14.1, BF.7.14.3, BF.7.14.4,BF.7.14.5,BF.7.14.6,accounted for 50.98%. BA.5.2 and its descendant lineages, including BA.5.2, BA.5.2.1, BA.5.2.6, BA.5.2.48, BA.5.2.49, BA.5.2.50, DZ.1, accounted for 47.71%. Other variants accounted for 1.31%, including BN.1.3, BN.1.5, BA.5.1, We further analyzed the trend of the proportion of variants over time. From December 2022 to February 2023, the proportion of BA.5.2 was higher than that of BF.7. But after February, BF.7 accounted for more than BA.5.2, reaching 80.95% at the highest point (March 11–17, 2023). However, the proportion of BF.7 began to decline in late March (Figure 4). Overall, the proportions of BF.7 and BA.5.2 alternately changed over time. The proportion of BF.7 rose rapidly from February to March, but began to decline at the end of March, while the proportion of BA.5.2 declined from February to March and rebounded at the end of March.

We conducted phylodynamic analysis on the BF.7 and BA.5.2 variants from December 2022 to March 2023.The maximum clade credit tree (MCC tree) results showed that both BF.7 and BA.5.2 exhibited a large amount of genetic diversities in December 2022.The results of Bayesian Skyline Plots showed that the effective population size of BF.7 and BA.5.2 both increased exponentially during this period. BA.5.2 increased sharply around December 13–December 23 and then entered a stable period. BF. 7 increased sharply around December 20–December 30 and then entered a stable period (Figure 5).