. 2021 Oct;2(10):e508–e517. doi: 10.1016/S2666-5247(21)00146-4

Table 1.

Characteristics of whole-blood RNA signatures for viral infection included in analysis

	Model^*	Discovery populations	Discovery settings	Discovery approach	Validation populations	Intended application
AndresTerre11²³	Geometric mean of all genes (influenza meta-signature)	Five cohorts of children and adults with influenza; adults challenged with influenza; and adults with bacterial pneumonia	UK, USA, and Australia	Differential expression followed by leave-one-cohort-out strategy and filtering for heterogeneity of effect size, using genome-wide data	Eight cohorts of children or adults with influenza or bacterial infection; adults challenged with influenza; and adults vaccinated against influenza	Influenza vs bacterial or other viral infection
Henrickson16²⁴	Difference in geometric means between upregulated and downregulated genes (influenza paediatric signature score)	Four cohorts of children with influenza-like illness	USA	Meta-analysis and leave-one-out strategy to identify common genes using genome-wide data	Two cohorts of children or adults with influenza	Influenza infection vs healthy
Herberg2¹⁰	Disease risk score^†	Children with viral or bacterial infection	UK, USA, and Spain	Elastic net followed by forward selection–partial least squares, using significantly differentially expressed transcripts	Children with bacterial or viral infection, inflammatory disease, or indeterminate diagnosis	Viral vs bacterial infection in febrile children
IFI44L¹⁴	NA	Children with viral or bacterial infection¹⁰	UK, USA, and Spain	Elastic net followed by forward selection–partial least squares, using significantly differentially expressed transcripts	Children with bacterial or viral infection	Viral vs bacterial infection in febrile children
IFIT3;RSAD2²² ^‡	NA	Three cohorts of adults challenged with rhinovirus, influenza, or RSV³⁵	UK and USA	Sparse latent factor regression analysis on genome-wide data³⁵ followed by regularised logistic regression on the resulting 30-gene signature	Close contacts of students with acute upper respiratory viral infections	Pre-symptomatic viral infection vs healthy
Lopez7¹⁵	Sum of weighted gene expression values (bacterial vs viral classifier)	Children and adults with viral, bacterial, or non-infectious acute respiratory illness¹⁹	USA	Support vector machine analysis using genome-wide data	Children with acute viral or bacterial infections³⁶	Viral vs bacterial respiratory infection
Lydon15¹¹	Logistic regression (viral classifier)^§	Adolescents and adults with viral, bacterial, or non-infectious acute respiratory illness	USA	LASSO regression analysis using 87 selected target genes from previously derived signatures¹⁹^,²¹	Patients with viral or bacterial co-infection or suspected bacterial infection	Viral vs bacterial respiratory infection
MX1³⁷	NA	NA	NA	Preselected due to biological plausibility	Adults challenged with the live yellow fever virus vaccine	Viral infection vs healthy
OLFM4²⁵	NA	Children with RSV infection	The Netherlands	Differential expression and prediction analysis of microarrays classifier training using genome-wide data	A second cohort of children with RSV infection	Severity of RSV infection in children
Pennisi2²⁰	Disease risk score^†	Children with viral or bacterial infection¹⁰	UK, USA, and Spain	Elastic net followed by forward selection–partial least squares, using significantly differentially expressed transcripts,¹⁰ then selection of an adequately expressed transcript for use in RT-LAMP	Children with bacterial or viral infection	Viral vs bacterial infection in children
Sampson10¹³	Disease risk score (combined SeptiCyte score)	Eight cohorts of neonates, children, and adults with bacterial infections	UK, USA, Estonia, and Australia	Regression analysis of transcript pairs using the 6000 most highly expressed genes from each dataset	Unselected consecutive patients presenting to the emergency department with febrile illness	Viral vs bacterial in febrile patients
Sampson4¹⁶	Disease risk score (Septicyte VIRUS)	Ten cohorts of children and adults with viral infections; two cohorts of adults challenged with influenza; and two cohorts of macaques challenged with Lassa virus or lymphocytic choriomeningitis virus	USA, Brazil, Finland, and Australia	Regression analysis of transcript pairs using the 6000 most highly expressed genes from each dataset	Seven human cohorts and six non-human mammal cohorts infected or challenged with viruses across all seven of the Baltimore virus classification groups	Viral vs non-viral conditions
Sweeney11¹⁷	Difference in geometric means between upregulated and downregulated genes, multiplied by ratio of counts of positive to negative genes (Sepsis metascore)	Nine cohorts of patients with sepsis or trauma	USA, Australia, Spain, Greece, the Netherlands, Norway, Canada, and UK	Greedy forward search of 82 differentially expressed genes identified by multicohort analysis	12 cohorts of adults with viral or bacterial sepsis, or trauma	Viral or bacterial sepsis vs sterile inflammation
Sweeney7¹²	Difference in geometric means between upregulated and downregulated genes, multiplied by ratio of counts of positive to negative genes (bacterial or viral metascore)	Eight cohorts of children and adults with viral and bacterial infections	USA, Australia, UK	Greedy forward search of 72 differentially expressed genes identified by multicohort analysis	24 cohorts of children and adults with viral or bacterial infections, or healthy controls	Viral vs bacterial infection
Trouillet-Assant6¹⁸	Median expression of six interferon-stimulated genes (interferon score³⁸)	NA	NA	Differential expression using 15 preselected interferon-stimulated genes	Febrile children with bacterial or viral infection	Viral vs bacterial infection in febrile children
Tsalik33¹⁹	Logistic regression (viral ARI classifier)^§	Children and adults with viral, bacterial, or non-infectious acute respiratory illness, and healthy controls	USA	LASSO regression analysis using the 40% of microarray probes with the largest variance after batch correction	Five cohorts of children or adults with viral, bacterial, or non-infectious respiratory illness, or viral or bacterial co-infection	Viral vs bacterial acute respiratory illness
Yu3;IFI27³⁹	Yu3: mean expression (non-RSV infections vs controls); IFI27: NA	Children with acute respiratory illness and a positive result for a viral infection on a nasopharyngeal swab	USA	Modified supervised principal component analysis using all expressed transcripts	Children with RSV or rhinovirus infection	Viral vs healthy in children
Zaas48²¹	Probit regression (viral classifier)^§	Two cohorts of adults challenged with influenza A H3N2 or H1N1	USA	Elastic net using 48 selected genes comprised of: 29 derived as a signature in a previous study,³⁵ seven shown to be downregulated in analysis of influenza challenge time course data,⁴⁰ and 12 control genes	Adults presenting to the emergency department with fever and healthy controls	Viral vs bacterial acute respiratory illness

Log₂-transformed transcripts per million data were used to calculate all signatures. NA=not applicable. RSV=respiratory syncytial virus. LASSO=least absolute shrinkage selector operator. RT-LAMP=reverse transcription loop-mediated isothermal amplification.

Where applicable, the name of the signature from the original publication is indicated in brackets.

^†

Defined as the sum of downregulated genes subtracted from the sum of upregulated genes.

^‡

Study by McClain and colleagues²² sought to validate a 36-transcript signature for the detection of respiratory viral infections. Model coefficients for the 36-transcript model are not provided; therefore, we included in this analysis the two best performing single transcripts from the study, since they had similar performance to the full model in the original publication.

^§

Logistic and probit regression models were calculated on the linear predictor scale using model coefficients from original publications.