each study, main features, genes/loci examined and summary of the main findings are reported (Continued) Author, year Population description Country Disease Primary outcome Other outcomes Sample N Cases N Controls N Notes on cohorts Gene/locus Conclusions Ref. Replication of previous study failed. Analysis of family data was done. Wang SF, 2011 Affected health care workers followed-up at the Dept. of Health, Taipei Taiwan SARS Manifest disease Prolonged neutralising antibody expression 97 56 41 HLA-B, HLA-Cw, HLA-DR HLA-Cw1502 and DR0301 possibly associated with resistance to SARS infection [54] Yuan FF, 2014 SARS patients admitted at Hong Kong Chinese University Hospital Hong Kong SARS Manifest disease Severity 18950 176 18774 48 cases classified as severe HLA-A, HLA-B, HLA-DR Negative results [55] Hajeer AH, 2016 Patients with laboratoryconfirmed MERSCoV infection admitted at King Abdulaziz Medical City, Riyadh Saudi Arabia MERS Manifest disease Severity 184 23 161 HLA-DRB1, HLA-DQB1 HLA-DRB1*11:01 nominally associated with severe MERS [56] Kuo CL, 2020 COVID-19 patients from the UN Biobank COVID19 COVID-19 positivity status 323570 622 322948 General population as control APOE APOE ε4 homozygous genotype associated, independent of preexisting dementia [57] Wang W, 2020 Donors of convalescent plasma who recovered from COVID-19 in Zhejiang, China China COVID19 Manifest disease 3630 82 3548 No severe case; 242 controls for HLA-DP1 locus HLA-A, HLA-B, HLA-C, HLA-DQ, HLA-DR, HLA-DP HLA haplotype C*07:29 and B*15:27 nominally associated with SARS-CoV-2 infection [58] Zhang Y, 2020 Patients with COVID-19 admitted to Beijing Youan Hospital, China China COVID19 Disease severity 80 80 24 severe cases vs 56 mild IFITM3 Association with severity [59] Notes and legend: gene and polymorphism names were reported as described in articles; the dbSNP nomenclature was added in brackets; statistics was reported if the significant association was found, not significant was annotated otherwise. ICU intensive care unit, LDH lactate dehydrogenase, WBC white blood counts, LR logistic regression, ns not significant, n/r not reported, n/a not applicable, #: Same cohort used by Itoyama S 2004, 2005, Keicho N 2009; § Same cohort used by Yuan FF 2005, 2007; $: Same cohort used by Chan VS 2006, Chan KYK 2007, 2010, Ching JCY 2010. Di Maria et al. Human Genomics (2020) 14:30 Page 8 of 19 All studies assumed 0.05 as the threshold of significance (as reported or inferred from article reading). Correction for multiple testing was not common in included studies, though hypothesis testing in genetic association studies implies multiple testing (see Table 3). One attempt to validate findings on an independent cohort was found [44]. Genetic association Table 1 reports all studies included. For each study, we reported genes/loci examined and a summary of main findings. Twenty-two articles reported the association analysis of CoV-related phenotypes with single-gene variants, either in coding or regulatory region. Ten genes were analysed in more than one study (Table 2), namely: ACE2, CLEC4M (L-SIGN), MBL, MxA (3 studies); ACE, CD209, FCER2, OAS-1, TLR4, TNF-α (2 studies). Out of these, only MBL and MxA provided positive signal of association in at least two studies. The CCL5/RANTES gene was reported in one article (Table 1) which investigated also an independent cohort of the same ancestry (i.e. Chinese) as a replication study; the second cohort provided a significant signal of association with severity but not with the manifest disease [44] (Table 3). Ten studies investigated the hypothesis of association with HLA loci (Tables 1 and 2). Three of these reported negative results: two independent studies analysed Chinese patients with SARS [48, 55]; one study [53] failed to replicate the association with HLA-B*0703 and HLADRB1*0301 previously reported [31]. Positive signals were found in 7 studies (5 on SARS, 1 in MERS and COVID-19). The following HLA haplotypes were found differentially distributed in patients with respect to the reference population: B*4601; B*0703; B*0301; Cw0801: DRB1*1202; Cw1502; DR0301; DRB1*11:01; C*07:29; B*15:27. None of these provided positive signals in more than one cohort (Table 3). Discussion This is the first comprehensive review which systematically collected all studies examining the involvement of human genetic variants in severe CoV infections, including COVID-19. Patarčić and co-workers had published the first extensive review including all studies investigating five common respiratory tract infectious diseases— pneumonia, tuberculosis, influenza, respiratory syncytial virus and SARS-CoV. At that time (last update: August 2015), they could cover the SARS-CoV-1 outbreak, though with some limitations such as the exclusion of HLA haplotypes [14]. A narrative review published well before the first SARS outbreak had pointed to CoV infection with a focus on animal studies [60]. By developing this comprehensive review, we recapitulated the body of knowledge about the influence of host germline genetic variants on the clinical phenotypes associated with CoVs infection. In total, we examined 32 articles that met the criteria and were eligible for data extraction. HLA haplotypes were examined in ten studies spanning the entire period from the SARS-CoV-1 outbreak