key: cord-0310203-ffdrsnst
authors: Rife Magalis, B.; Rich, S.; Tagliamonte, M. S.; Mavian, C.; Cash, M. N.; Riva, A.; Marini, S.; Moraga Amador, D.; Zhang, Y.; Shapiro, J.; Horine, A.; Starostik, P.; Pieretti, M.; Vega, S.; Lacombe, A. P.; Salinas, J.; STEVENSON, M.; Myers, P.; Morris, J. G.; Lauzardo, M.; Prosperi, M.; Salemi, M.
title: SARS-CoV-2 Delta vaccine breakthrough transmissibility in Alachua, Florida
date: 2021-11-11
journal: nan
DOI: 10.1101/2021.11.10.21266134
sha: beb850f5dbff6f83244cead429a021c59abcb6ce
doc_id: 310203
cord_uid: ffdrsnst

Background SARS-CoV-2 Delta variant has caused a dramatic resurgence in infections in the United Sates, raising questions regarding potential transmissibility among vaccinated individuals. Methods Between October 2020 and July 2021, we sequenced 4,439 SARS-CoV-2 full genomes, 23% of all known infections in Alachua County, Florida, including 109 vaccine breakthrough cases. Univariate and multivariate regression analyses were conducted to evaluate associations between viral load (VL) level and patient characteristics. Contact tracing and phylogenetic analysis were used to investigate direct transmissions involving vaccinated individuals. Results The majority of breakthrough sequences with lineage assignment were classified as Delta variants (74.6%) and occurred, on average, about three months (104 +- 57.5 days) after full vaccination, at the same time (June-July 2021) of Delta variant exponential spread within the county. Six Delta variant transmission pairs between fully vaccinated individuals were identified through contact tracing, three of which were confirmed by phylogenetic analysis. Delta breakthroughs exhibited broad VL values during acute infection (IQR 1.2-8.64 Log copies/ml), on average 38% lower than matched unvaccinated patients (3.29-10.81 Log copies/ml, p<0.00001). Nevertheless, 49-50% of all breakthroughs, and 56-60% of Delta-infected breakthroughs exhibited VL above the transmissibility threshold (4 Log copies/ml) irrespective of time post vaccination. Conclusions Delta infection transmissibility and general VL patterns in vaccinated individuals suggest limited levels of sterilizing immunity that need to be considered by public health policies. In particular, ongoing evaluation of vaccine boosters should address whether extra vaccine doses might curb breakthrough contribution to epidemic spread.

and level of evolutionary divergence from the original strain, have raised questions regarding the extent of protection of currently implemented vaccines against infection. [3] [4] [5] [6] [7] Lineage B.1.1.7 (Pango nomenclature 8, 9 ) , otherwise known as Alpha (WHO nomenclature 10 ), was detected in September 2020 in Kent, UK, and was estimated to have arrived in the US approximately two months later. 11 Other VOCs -Beta, Gamma, Delta and Epsilon -were later identified and have been circulating, at different frequencies, in different areas of the world. 3 14 Delta carries more than a dozen mutations, including L452R, also found in the California variant Epsilon, 15 associated with increased Spike stability and viral fusogenicity, which results in enhanced viral replication and infectivity, 16 as well as impaired immune response through antibody neutralization. 17 As of November 03, 2021, 66.9% of Americans were fully vaccinated, 18 yet emergence of the Delta variant, which has rapidly been spreading in the US since June 2021, has caused a dramatic resurgence of infections and hospitalizations. [19] [20] [21] Approximately 74% of infections with the Delta variant are followed by symptomatic onset 22 and are characterized by high viral load (VL) linked to higher transmission rates, 23 explaining how this variant has outpaced other VOCs and become the predominant variant in several countries, mostly driven by dissemination among the unvaccinated. 24 According to clinical trial results, a two-dose regimen of Pfizer-BioNTech (BNT162b2) vaccine confers 94.6% protection against severe disease, 6 while the Moderna (mRNA-1273) vaccine confers 94.1%. 25 . While vaccination continues to provide excellent protection against hospitalization and death, over the past six months there has been a gradual decline in vaccine efficacy against infection. 26 In the US, these vaccine breakthrough cases linked to VOCs have been reported since January 2021. 27 In July 2021, following multiple large public events in a Barnstable County, Massachusetts, town, an outbreak (90% Delta) of infection (albeit mild disease) was identified among Massachusetts residents traveling recently to the town, primarily comprised of fully vaccinated persons. 19 Within the same month, a similar report of widespread Delta circulation (95% of sequenced samples by July 24) among individuals in Wisconsin (USA) was published in medRxiv. 28 Both studies reported similar VL among vaccinated and unvaccinated individuals, which were estimated by using the number of polymerase chain reaction cycles required to quantify viral genetic sequence fragments (CT value) in a specimen. Previous work has shown that infectious SARS-CoV-2 can usually be recovered from specimens with CT < 25 -30 . 29 Yet, CT values are only a proxy for the level of virus shed in the nasal passages and, while these data suggest vaccinated individuals are susceptible to infection by the Delta variant and harbor the level of infectious virus required for further transmission, definitive evidence of this transmission has not been presented.

Moreover, CT values do not allow for a quantitative evaluation of the VL threshold required for potential transmission within a specific population, e.g., the fraction of vaccine breakthrough cases harboring a VL compatible with secondary transmissions, which both empirical data and theoretical studies have shown to be 4 Log copies/ml. 30 

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The Alachua County Department of Health, with assistance from designated UF personnel, has been responsible for contact tracing efforts for University of Florida (UF) students, faculty, staff, and other UFaffiliated people, including the UF Health Academic Medical Center (~123,000 total UF Affiliates) since the beginning of the epidemic. SARS-CoV-2 positive samples were collected for virus full genome sequencing, as part of this program, between October 2020 and August 2021 from patients hospitalized at UFHealth Shands Hospital Gainesville during this time period, as well as from testing sites of UF PathLabs in Gainesville serving Alachua County residents (Supplementary Table S1 ). Samples from positive patients in the Tampa Bay are and Miami Dade (provided by BayCare and University of Miami, respectively), which had been collected for other studies, were included in the sequence analysis for comparison purposes (Table S1 ). Full epidemiological investigations were conducted on Alachua County positive cases to collect exposure information, trace contacts, and provide disease transmission education.

Vaccine breakthrough cases were defined as individuals who were PCR-positive for SARS-CoV-2 and ≥14 days after the second dose of Pfizer or Moderna or first dose of Janssen/J&J. Repeat saliva samples were collected from possible breakthrough cases for sequencing. Patient samples and linked data were fully deidentified before sample processing. The study was reviewed and approved under the category of Public Health practice by the University of Florida Institutional Review Board (IRB) and the Florida Department of Health IRB.

Viral RNA, from either 200 µL of viral transport medium for NP swabs or 180 µL of saliva, was extracted for each sample using the QIAamp 96 Viral RNA Kit with the QIAcube HT (Qiagen, Germantown, MD) using the following settings with a filter plate: the lysed sample was premixed 8 times before subjecting to vacuum for 5 minutes at 25kP and vacuum for 3 min at 70kPa. Following 3 washes using the same vacuum conditions above, the samples were eluted in 100 µL AVE buffer followed by a . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

For each in-house sequencing run, sequences within the GISAID database associated with the state of Florida were extracted up to August 03, 2021, and added to the collection of high-coverage in-houseproduced genome sequences (N=3,110 from Alachua County, N=126 from Fort Myers, N=791, from Miami Dade), totaling 20,117 sequences, ranging from February 28, 2020, to August 03, 2021. Each Floridian sequence within this concatenated dataset was then used in a local alignment (BLAST) 33 search for the most (genetically) similar non-Floridian sequence in the GISAID database as of the current date and linked to two reference sequences including the best match (highest E-value) with a date occurring within one month following, as well as one month prior to, the sampling date of the Floridian sequence. 34 After removing duplicate sequences (sequences with same GISAID ID), the final GISAID reference dataset totaled 5,074, ranging from February 21, 2020, to July 27, 2021. Sequences were aligned in viralMSA 35 35 using the MN908947 reference sequence. Mutations potentially associated with contamination, recurrent sequencing errors, or hypermutability were masked using a VCF filter provided at https://virological.org/t/masking-strategies-for-sars-cov-2-alignments/480.

Following the first in-house sequencing run and first round of BLAST protocol described above, the multiple sequence alignment for all sequences up to November 25, 2020, was used in the reconstruction . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. of an initial maximum likelihood tree in IQ-TREE2 using the best-fit evolutionary model according to Bayesian Information Criterion 36 . 36 . The tree was updated with each round of sequencing, concatenation, and local alignment using phylogenetic placement in UShER. 37 The tree topology was also refined after each round using two rounds of subtree pruning and regrafting (maximum length of each move of 100) in FastTreeMP. 38 . Branch lengths were re-optimized using the gamma distribution. Lineages for all sequences were determined using the PangoLEARN model (updated daily), which is trained using GISAID SARS-CoV-2 sequences to classify incoming sequences based on molecular and epidemiological criteria. 8,9

The subtree containing sequences classified as Delta was pruned from the full tree for transmission cluster analysis. Root-to-tip regression was used to identify outliers, defined as sequences with extreme deviation in genetic divergence from the best-fit root of the tree, suggestive of incorrectly reported sampling dates, though true variation owing to natural processes such as recombination and positive selection cannot be ruled out. 39 . Two sequences were identified as outliers and pruned from the tree prior to downstream analysis using the ape package in R. 40 Transmission cluster identification within the resulting pruned tree was performed using a depth-first search among nodes restricted to minimal evolution (branch lengths) and sampling time difference (available from https://github.com/ProsperiLab/Dynamite). This transmission cluster identification approach is similar in nature to other phylogenetic cluster-picking algorithms (e.g., Phylopart 41 Phylopart 41 , ClusterPicker 42 ), with the exception that clusters were not restricted to monophyletic clades and an informed serial time interval was used to accept or reject sampled sequences based on a reasonable transmission window. As with other phylogenetic algorithms, a genetic distance threshold was required as criterion to accept or reject nodes into each cluster. Similar to Phylopart, the optimal threshold was chosen which maximized the number of clusters identified-4.23 x 10 -5 substitutions/site. Phylopart was also used for robust comparison, which resulted in one less cluster than the described algorithm for a range of optimum thresholds (3.86 x 10 -5 -5.03 x 10 -5 ). For the . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint described depth-first algorithm, individual nodes (either external or internal) were accepted as part of a cluster if the mean branch length of this node and all preceding accepted nodes did not exceed the optimal threshold. The serial time interval used as an additional criterion to accept tips into individual clusters was 6 days, based on meta-analysis performed by Rai et al. for SARS-CoV-2. 43 Similar to the branch length procedure, external nodes were accepted as part of a cluster if the above criterion was met and if the mean pairwise sampling time difference among the tip in question and previously accepted tips did not exceed the specified serial time interval threshold.

Levels of SARS-CoV-2 were determined using the 2019-nCoV_N1 assay (primer and probe set) with 2019-nCoV_N_positive control (IDT, Coralville, Iowa). Viral RNA was extracted as previously described then subjected to first strand synthesis using ProtoScript II Reverse Transcriptase according to the manufacturer's instructions (NEB, Ipswich, MA). Quantitative PCR was performed using TaqMan Fast

Advanced Master Mix (Waltham, MA) according to the manufacturer's instructions. A standard curve was generated using N1 quantitative standards 10-fold diluted to determine viral copies. The assay was run in triplicate including one non-template control.

We performed a linear regression on the VL as dependent variable, with vaccination status (yes/no), Delta lineage (yes/no, age (<21, 21-28, 28-44, >44 years old or unknown), gender (male, female, unknown), and days from January 2021 as covariates. We also considered an interaction term between vaccination status and Delta, and coefficients were calculated with and without the inclusion of the interaction term.

For interpretation of individual coefficients, we can assume that either vaccination status or Delta variant might be confounded by age, gender and time passed, and give a possibly causal interpretation of said coefficients as total effects under that adjustment set. Univariable and multiple regression analyses were performed to evaluate associations between patient characteristics and either VL or RT-PCR cycle threshold (CT) levels. The analyses considered demographic variables [age (years), sex (male vs. female), . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint race (white vs. non-white), ethnicity (Hispanic vs. non-Hispanic)] and clinical information [vaccine brand (Moderna, Pfizer, or J&J), symptom status (yes/no)], and time between vaccination and onset (days). To investigate associations between these variables and levels of infectivity, outcomes were also categorized as binary variables for multivariable logistic regression analysis. Patients were considered above the transmissibility threshold (i.e., potentially infectious) if they had a VL level greater than or equivalent to Log 4 copies per ml of sample. 31

Between October, 2020, and first week of August, 2021, we sequenced 4,439 SARS-CoV-2 samples from patients in Alachua County (Table S1) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint that have shown the effectiveness of currently available vaccines. However, while the spike in breakthrough cases during the month of July did coincide with a significant vaccination scale up in Alachua County 18 , the majority (71%) of these patients had already been fully vaccinated for more than 90 days ( Figure 1C ) and became infected by the Delta variant in July, while its frequency was exponentially increasing among the unvaccinated population ( Figure 1B) , indicating that the increase in breakthroughs was indeed linked with the emergence of the Delta variant. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint Contrary to other reports 19, 28 , however, Delta-infected breakthrough cases in Alachua County exhibited an average 38% VL reduction compared to unvaccinated Delta cases (p<0.00001, two tailed U-test, after

Bonferroni correction), and 34% (p<0.00001) compared to unvaccinated non-Delta cases ( Figure 2 ). The multivariable analysis did not find any association between VL and age, gender, race, ethnicity, or vaccine type in the sample (Table S2 ). Yet, Delta infections had a strong association with VL above the transmissibility threshold of Log 4 RNA copies/ml, 30,31 with a 2.46 odd ratio (OR) and confidence interval (CI) 1.05 -5.97 in the unadjusted analysis, and OR 3.04 and CI 1. 16 -8.54 in the adjusted analysis (Table S3 ). In particular, the majority of vaccine breakthrough cases infected with the Delta variant (58.5%) exhibited a VL above the required threshold for potential transmission (Figure 2 ). is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint intermixed throughout the phylogeny, clusters were primarily comprised of local transmissions ( Figure   3A ). Vaccination breakthroughs were also intermixed throughout the phylogeny; however, all contact tracing-identified transmission pairs belonged to cluster c3 ( Figure 3B (Figure 3) . Therefore, the three donor individuals were removed prior to regression analysis, leaving the three recipient individuals which we know to be unrelated via direct transmission.

Remaining Delta vaccination breakthroughs with VL data were sparsely placed throughout the tree (45 in background population and remaining two in clusters c2 and c6) and could thus be included. Only one unvaccinated Delta individual with VL data was observed within a transmission cluster (c2) and was considered to be distinct from the vaccinated individual in this cluster by significant branch support and so was included in regression analysis. Non-Delta sequences were not included in the phylogeny; however, samples from this population were chosen for VL analysis so that a wide range of collection times and variants were included, minimizing the probability of epidemiological linkage and the need for phylogenetic assessment of independence. Multiple regression analysis had already found no evidence of . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint association between VL and time since full vaccination (8 -186 days) , time interval between symptom onset and sampling, or vaccine type using either linear or logistic methods (Supplementary Table S2 ).

There was also no correlation between distribution of VL over time elapsed since full vaccination, defined Figure S2) .

The rapid emergence, in July 2021, of Delta infections correlated with a major spike in reported cases, hospitalizations, and deaths in Florida and in Alachua County, due primarily to infections in unvaccinated persons 44 . It also coincided with a sudden spike in vaccine breakthrough cases, demonstrating that direct transmission from vaccine breakthroughs can occur, with more than half of the Delta breakthroughs in our sample harboring sufficient viral copies to transmit the virus during acute infection.

Overall, the number of breakthrough cases detected, during our nearly nine-month genomic epidemiology surveillance program, represented approximately 2% of the sequenced cases, a low number of vaccine failures in agreement with the known effectiveness of the vaccines according to randomized clinical trials. Even assuming that we may have missed a substantial number of vaccine breakthrough . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint infections in asymptomatic individuals, we have no records of severe/fatal COVID-19 cases that involved vaccinated patients between December 2020 (when vaccination began in Alachua County) and end of July 2021. Our VL data also show that, despite Delta variant high transmissibility -as evidenced by its ability to become the majority variant within a month since its emergence in the county -vaccination is associated with lower viral replication in breakthrough cases compared to unvaccinated patients. While other studies have reported similar VL among vaccinated and unvaccinated individuals 19, 28 , such studies utilized CT values as a proxy for VL among groups of patients rather than the actual number of RNA copies, which may in part explain the discrepancy. Moreover, studies such as the one in Barnstable is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ;  transmissible/pathogenic or vaccine-resistant strains. Therefore, while additional observations based on larger numbers of patients are necessary, our work indicate that the differential level of sterilizing immunity, or lack of thereof, that may be present in the vaccinated population, is an important factor to be considered for the implementation of the next phase of vaccination and intervention policies. In particular, the ongoing evaluation of vaccine boosters 47, 48 should address, besides effectiveness against disease, whether the administration of extra vaccine doses may reduce the number of breakthrough cases, or at least decrease the proportion of breakthroughs exhibiting VL above the transmissibility threshold during the acute phase of infection, thus improving sterilizing immunity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. For each group pair, a two-tailed Mann-Whitney U test was executed. P-value and effect size are shown on top for those comparisons between groups that were significant at the 5% level after Bonferroni correction for multiple tests. days after symptoms onset (see Table 1 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint Log RNA copies/ml Days from full vaccination 48.8% 50% 55.6% 60%

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 11, 2021. ; https://doi.org/10.1101/2021.11.10.21266134 doi: medRxiv preprint

SARS-CoV-2 Variant Classifications and Definitions

Preliminary genomic characterisation of an emergent SARS-CoV-2 lineage in the UK defined by a novel set of spike mutations

Efficacy of the ChAdOx1 nCoV-19 Covid-19 Vaccine against the B.1.351 Variant

Three-quarters attack rate of SARS-CoV-2 in the Brazilian Amazon during a largely unmitigated epidemic

Long-term intermittent detection of SARS CoV 2 in the upper respiratory tract: what is the meaning of it?

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Effectiveness of Covid-19 Vaccines against the B.1.617.2 (Delta) Variant

A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology

Addendum: A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology

WHO. Tracking SARS-CoV-2 Variants

SARS-CoV-2 variants of concern and variants under investigation in England

Detection of a SARS-CoV-2 variant of concern in South Africa

Rapid emergence and spread of SARS-CoV-2 gamma (P.1) variant in Haiti

What scientists know about new, fast-spreading coronavirus variants

Emergence of a Novel SARS-CoV-2 Variant in Southern California

SARS-CoV-2 spike L452R variant evades cellular immunity and increases infectivity

Complete map of SARS-CoV-2 RBD mutations that escape the monoclonal antibody LY-CoV555 and its cocktail with LY-CoV016. bioRxiv 2021

COVID-19 Vaccinations in the United States

Outbreak of SARS-CoV-2 Infections, Including COVID-19 Vaccine Breakthrough Infections, Associated with Large Public Gatherings

SARS-CoV-2 B.1.617.2 (Delta) Variant COVID-19 Outbreak Associated with a Gymnastics Facility -Oklahoma

Transmission event of SARS-CoV-2 delta variant reveals multiple vaccine breakthrough infections

Transmission dynamics and epidemiological characteristics of Delta variant infections in China

Transmission Dynamics of an Outbreak of the COVID-19

Delta Variant B.1.617.2 -Guangdong Province, China

Delta's rise is fuelled by rampant spread from people who feel fine

Efficacy and Safety of the mRNA-1273 SARS-CoV-2

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine through 6 Months

COVID-19 Vaccine Breakthrough Infections Reported to CDC -United States

Despite Vaccination when the Delta Variant is Prevalent -Wisconsin

Duration of infectiousness and correlation with RT-PCR cycle threshold values in cases of COVID-19

Viral load and contact heterogeneity predict SARS-CoV-2 transmission and super-spreading events

Transmission of COVID-19 in 282 clusters in Catalonia, Spain: a cohort study

Comparison of two highly-effective mRNA vaccines for COVID-19 during periods of Alpha and Delta variant prevalence. medRxiv 2021

Basic local alignment search tool

SARS-CoV-2 shifting transmission dynamics and hidden reservoirs potentially limit efficacy of public health interventions in Italy

Massively scalable reference-guided multiple sequence alignment of viral genomes

IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era

Ultrafast Sample placement on Existing tRees (UShER) enables real-time phylogenetics for the SARS-CoV-2 pandemic

FastTree 2--approximately maximum-likelihood trees for large alignments

Exploring the temporal structure of heterochronous sequences using TempEst (formerly Path-O-Gen)

ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R

QuRe: software for viral quasispecies reconstruction from nextgeneration sequencing data

Automated analysis of phylogenetic clusters

Estimates of serial interval for COVID-19: A systematic review and meta-analysis

COVID-19 Case Surveillance Public Use Data with Geography

The phylogenetic regression