Research Article

Evolutionary Biology

Viral genome sequence datasets display pervasive evidence of strand-specific substitution biases that are best described using non-reversible nucleotide substitution models

Division of Computational Biology, Institute of Infectious Diseases and Molecular Medicine, Department of Integrative Biomedical Sciences, Faculty of Health Sciences, University of Cape Town, South Africa
Department of Human Biology, Faculty of Health Sciences, University of Cape Town, South Africa
Centre for Infectious Disease and Epidemiology Research, School of Public Health and Family Medicine, University of Cape Town, South Africa
Centre for Biomedical Engineering, School of Engineering, Brown University, United States
Institute for Genomics and Evolutionary Medicine, Department of Biology, Temple University, United States
Wellcome Center for Infectious Diseases Research in Africa, Institute of Infectious Disease and Molecular Medicine and Department of Medicine, University of Cape Town, South Africa

Sep 30, 2025

https://doi.org/10.7554/eLife.87361.3

Open access
Copyright information

Figures
Tables
Additional files

3 figures, 5 tables and 2 additional files

Figures

Figure 1

Download asset Open asset

Ternary plots illustrating the relative fit of the NREV12, NREV6, and GTR nucleotide substitution models based on weighted small sample corrected Akaike information criterion (AIC-c) scores for 30 dsDNA, 31 dsRNA, 33 ssDNA, and 47 ssRNA virus nucleotide sequence datasets.

These plots were produced using the Akaike weights function with an overlaid density function (implemented in the qpcR package of RStudio; Ritz and Spiess, 2008) to indicate point densities. Each model is represented by a corner of the triangles, and each circle represents the relative fit of each of the three models to a single nucleotide sequence dataset. The sides of the triangle represent model support axes ranging from 0% to 100%, with the position of a circle in relation to each of the sides of the triangle indicating the probability of models best describing the nucleotide sequence dataset that is represented by that point. Red colours represent a very high density of nucleotide sequence datasets that favour a particular model, blue colours indicate a lower, but still substantial, density of datasets that favour a particular model.

Figure 2

Download asset Open asset

Weighted Robinson-Foulds distances between inferred and true phylogenetic trees for datasets simulated with different degrees of nucleotide substitution non-reversibility and different average pairwise sequence identities (APIs) (~75%, ~80%, ~85%, ~90%, and ~95%).

‘ns’ above a pair of box and whisker plots indicates a paired t-test adjusted p-value of ≥0.05 and ‘*’ indicates a paired t-test adjusted p-value of <0.05.

Figure 3

Download asset Open asset

Phylogenetic tree inferred from an alignment of real sequences (Avian Leukosis virus) that was used to simulate datasets with degrees of non-reversibility (DNRs) varying from 0 to 20.

The alignment of Avian Leukosis virus had an average sequence identity (API) of ~90%, and the branches of this tree were scaled to produce four other trees reflecting branch tip sequences with approximate pairwise identities of ~75%, ~80%, ~85%, and~95%.

Tables

Table 1

Akaike information criterion (AIC) scores and likelihood ratio test (LRT) results for double-stranded DNA virus datasets.

The lowest small sample corrected AIC (AIC-c) scores indicating the best fitting models are in bold.

Virus family	Dataset	AIC score GTR	AIC score NREV-6	AIC score NREV-12	p-Value GTR vs NREV-12	p-Value NREV-6 vs NREV-12	DNR
Papillomaviridae	APPV 6	35099.5	35108.0	35102.2	>0.05	0.007	0.089
	HPV18_2	25202.9	25174.6	25179.2	<0.001	>0.05	0.323
	HPV45_2	23600.6	23599.0	23602.9	>0.05	>0.05	0.285
	HPV16_2	29734.0	29664.5	29665.4	<0.001	>0.05	0.371
	HPV31	24681.4	24677.3	24672.8	0.002	0.01	0.165
	HPV6_1	31199.1	31150.0	31141.2	<0.001	<0.001	0.451
	LPV	67165.7	67188.1	67145.5	<0.001	<0.001	0.42
	DPV	69829.7	69889.2	69835.1	>0.05	<0.001	0.056
	XPV	95455.6	95617.1	95452.2	<0.001	<0.001	0.072
	BATV	134821	134511	133322	<0.001	<0.001	0.402
Polyomaviridae	JC_2	51806.7	51819.6	51812.0	>0.05	0.003	0.089
	BK_2	21472.6	21472.7	21471.1	0.03	0.03	0.244
	SV40	16859.8	16858.0	16858.4	0.037	>0.05	0.567
	BPV	148614.9	148573.8	148585.2	<0.001	>0.05	0.064
Caulimoviridae	CMV	124083.9	124221.0	123888.6	<0.001	<0.001	0.351
	CSSV	145327.0	146575	145202	<0.001	<0.001	0.158
	SVBV	138575	138488.1	138464.7	<0.001	<0.001	0.174
	DBAV	46495.5	46514.1	46502.0	>0.05	<0.001	0.0335
	RTBV	54987.9	55350.1	54991	>0.05	<0.001	0.082
	BDV	376325.2	376647.6	376029.9	<0.001	<0.001	0.140
Siphoviridae	CLV	237362.3	237351.8	237348.6	<0.001	<0.01	0.070
Tectiviridae	TTIV	913864.9	913915.4	913773.1	<0.001	<0.001	0.279
Adenoviridae	FAV_C	3074086.7	3074207.5	3073739.1	<0.001	<0.001	0.169
	FAV_E	103482.3	103222.7	102636.7	<0.001	<0.001	0.357
	FAV_D	2326925.6	2325719.4	2324784.5	<0.001	<0.001	0.551
	FAV_A	705328.5	705436.5	705197.8	<0.001	<0.001	0.645
	HMAV_B	103796.7	103937.44	103753.8	<0.001	<0.001	10.890
	HMAV_D	1748635.2	1749769	1748119.1	<0.001	<0.001	0.646
	HMAV_C	2851144.5	2851357.1	2851133	0.006	<0.001	0.0225
	HMAV_E	1915044.8	1915065.3	1914998	<0.001	<0.001	0.049

Table 2

Akaike information criterion (AIC) scores and likelihood ratio test (LRT) results for double-stranded RNA datasets.

The lowest small sample corrected AIC (AIC-c) scores indicating the best fitting models are in bold.

Virus family	Dataset	AIC score GTR	AIC score NREV-6	AIC score NREV-12	GTR vs NREV-12	NREV-6 vs NREV-12	DNR
Birnaviridae	AQBV	31754.9	31853.3	31721.9	<0.001	<0.001	0.219
	GBV_A	47176.9	47347.2	47154.8	<0.001	<0.001	0.142
	IPNV	79186.2	79221.9	79182.4	0.0145	<0.001	0.162
	GBV_B	39313.7	39062.8	38938.7	<0.001	<0.001	0.201
Reoviridae	BTV_A	34803.5	34895.1	34801.3	0.03	<0.001	0.042
	BTV_B	48849.9	48893.	48837.1	<0.001	<0.001	0.043
	BTV_C	28350.9	28386.5	28350.8	>0.05	<0.001	0.061
	BTV_D	24969.1	24947.3	24894.0	<0.001	<0.001	0.191
	BTV_F	20622.7	20708.5	20610.2	<0.001	<0.001	0.067
	BTV_G	63349.9	63485.0	63345.9	0.00426	<0.001	0.040
	BTV_H	20596.7	20685.5	20586.1	<0.001	<0.001	0.118
	BTV_I	17592.7	17622.5	17588.8	0.01	<0.001	0.095
	BRVA_C	41206.7	41187.4	41137.1	<0.001	<0.001	0.128
	HRVA_A	17030.5	17043.2	17035.5	>0.05	0.003	0.036
	HRVA_B	8275.1	8280.3	8281.7	>0.05	>0.05	0.087
	HRVA_C	12815.1	12842.6	12807.6	0.003	<0.001	0.132
	HRVA_D2	8036.8	8041.0	8043.7	>0.05	>0.05	0.057
	HRVA_E	7045.9	7056.1	7053.3	>0.05	0.02	0.102
	HRVA_F	7046.0	7056.7	7053.4	>0.05	0.02	0.0710
	HRVA_G	18424.2	18434.0	18425.1	>0.05	<0.001	0.123
	HRVA_H	20431.4	20413.87	20420.5.6	0.002	>0.05	0.163
	PRVA_A	28540.7	28441.9	28398.7	<0.001	<0.001	0.204
	PRVA_B	14757.7	14775.5	14732.6	<0.001	<0.001	0.351
	HRVC_A	6713.2	6718.2	6712.3	0.045	0.007	0.124
	PTOV	202011.3	202106.5	201878.5	<0.001	<0.001	0.039
	FJV_B	9274.1	9250.0	9250.9	<0.001	>0.05	0.194
Endornaviridae	EDV	1771992.8	1772689.1	1771950.6	<0.001	<0.001	0.121
Endornaviridae	BPAV	70386.5	70540.2	70390.7	>0.05	0.00	0.047
Totiviridae	TTV	617302.6	617462.6	617172.9	<0.001	<0.001	0.052
Totiviridae	GDV	80435.8	80396.5	80387.7	<0.001	0.002	0.109
Hypoviridae	HPV	66859.8	66899.8	66857.8	0.03	<0.001	0.057

Table 3

Small sample corrected Akaike information criterion (AIC-c) scores and likelihood ratio test (LRT) results for single-stranded DNA datasets.

The lowest AIC-c scores indicating the best fitting models are in bold.

Virus family	Dataset	AIC score GTR	AIC score NREV-6	AIC score NREV-12	p-Value GTR vs NREV-12	p-Value NREV-6 vs NREV-12	DNR
Nanoviridae	BBTV M	15044.3	15207.9	14984.4	<0.001	<0.001	0.662
	BBTV N	10605.6	10686.2	10595.2	<0.001	<0.001	0.533
	BBTV R	18484.5	18544	18480.8	>0.05	<0.001	0.609
	BBTV S	12718.9	12757.2	12707.3	<0.001	<0.001	0.728
	CCDV	38622.7	38632.0	38630.5	>0.05	0.03	0.050
	MDV	36232.8	36063	36064	<0.001	>0.05	0.142
	PYDV	56138.4	56076.6	56056.4	<0.001	<0.001	0.187
	FBNS	100153.6	100135.6	100120.5	<0.001	<0.001	0.098
Geminiviridae	Begomo 5	28192.1	28311.9	28192.5	>0.05	<0.001	0.1995
	Begomo 6	16743.0	16722.6	16724.1	<0.001	>0.05	0.214
	Begomo 9	8517.6	8540.8	8515.6	0.03	<0.001	0.312
	Dicot 1	44730.7	44594.3	44583.3	<0.001	<0.001	0.200
	Dicot 2	39909.9	39919.8	39917.9	>0.05	<0.001	0.100
	MSV	252645.3	254347.5	254347.5	<0.001	<0.001	0.144
	PanSV	94601.2	94600.3	94593.7	<0.001	<0.001	0.182
	WDV	35301.7	35313.2	35253.8	<0.001	<0.001	0.1033
Circoviridae	BFDV	17256.7	17262.7	17246.7	<0.001	<0.001	0.224
	DG_CV	12754.8	12779.5	12758.3	>0.05	<0.001	0.116
	PiCV	19180.5	19192.5	19191.0	>0.05	0.04	0.117
	CCCC	84435.7	84377.4	84315.3	<0.001	<0.001	0.132
	BTC	262910.4	262060.1	261985.4	<0.001	<0.001	0.178
	POCV2	24940.9	24953.8	24915.8	<0.001	<0.001	0.162
	CCV	90307.9	90301.5	90285.9	<0.001	<0.001	0.114
Anelloviridae	TTV_1	825811	826800	825292	<0.001	<0.001	0.513
Anelloviridae	TTSV	332287.9	332397.4	332258.2	<0.001	<0.001	1.560
Parvoviridae	MVM	26756.3	26743.9	26686.9	<0.001	<0.001	0.148
	HPV	67051.2	67080.1	67001.8	<0.001	<0.001	0.235
	CPV	85731	85695	85689.3	<0.001	0.007	0.062
	PPV	163006.8	163090.7	162995.9	<0.001	<0.001	0.143
	CAV_P	37073.3	37115.5	37065.7	<0.001	<0.001	0.162
Microviridae	BMV	31175.3	31164.8	31147.3	<0.001	<0.001	0.188
Pleolipoviridae	APV	85700.2	85617.4	85402.8	<0.001	<0.001	0.204
Pleolipoviridae	BPV	204797.5	204802.3	204796.7	0.04	0.007	0.064

Table 4

Small sample corrected Akaike information criterion (AIC-c) scores and likelihood ratio test (LRT) results for single-stranded RNA datasets.

The lowest AIC-c scores indicating the best fitting models are in bold.

Virus family	Dataset	AIC score GTR	AIC score NREV-6	AIC score NREV-12	p-Value GTR vs NREV-12	p-Value NREV-6 vs NREV-12	DNR
Astroviridae	HAV	94580.7	94926.3	94548.1	<0.001	<0.001	0.096
	BAV	188307.1	188572.9	188144.9	<0.001	<0.001	0.108
	MMV	281072.2	281094.5	281076.9	>0.05	<0.001	0.072
	PAV	150626.88	150827.6	150609.5	<0.001	<0.001	0.069
	CKV	90902.3	91233.1	90873.0	<0.001	<0.001	0.083
	GA	64998.5	65223.9	64975.9	<0.001	<0.001	0.110
	CAV_A	85558.8	85617.4	85547.3	<0.01	<0.001	0.076
Bromoviridae	CMV RNA1	34197.5	34198.8	34147.7	<0.001	<0.001	0.124
	CMV RNA2	31398.2	31455.9	31388.7	<0.001	<0.001	0.091
	CMV RNA3	24337.2	24360.3	24343.9	>0.05	<0.001	0.073
	AMS	24337.2	24360.3	24343.9	>0.05	<0.001	0.073
	PSV	67707	67786.5	67691	<0.001	<0.001	0.048
Caliciviridae	LAV	73042.8	73102.4	72984.6	<0.001	<0.001	0.120
	NoV	207667.2	207777.5	207660	<0.001	<0.001	0.047
	VSV	235936.4	236051.4	235913.3	<0.001	<0.001	0.046
Closteroviridae	CTV	30062.2	29980.4	29960.1	<0.001	<0.001	0.272
Flaviviridae	DGV_T1	69771.9	70030.5	69776.2	>0.05	<0.001	0.063
Flaviviridae	JEV	146920.8	148101.5	146885.5	<0.001	<0.001	0.091
Hepeviridae	HPVE1	200439.5	200863.8	200179.8	<0.001	<0.001	0.073
Hepeviridae	HPVE2	155709.1	155983.8	155518.6	<0.001	<0.001	0.088
Picornaviridae	ENV_A	552287.9	553535.5	551794.1	<0.001	<0.001	0.061
	HRV_A	102218.7	102267.0	101550.7	<0.001	<0.001	0.285
	AIV	101073.1	101136.7	101052.2	<0.001	<0.001	0.093
	AHP	139635.7	140119.6	139506.9	<0.001	<0.001	0.170
	ECV	82078.9	82181.0	82065.8	<0.001	<0.001	0.066
	CDV	130551.3	130896.7	130478.3	<0.001	<0.001	0.086
	TCV	53027.3	53029	53023	0.0151	0.0422	0.033
	FMDV	455180.6	455582.6	454806.1	<0.001	<0.001	0.117
Fusariviridae	FRV	52413.1	52470.6	52418.4	>0.05	<0.001	0.076
Retroviridae	HIV1_setA	344014.4	344295.1	343669.7	<0.001	<0.001	0.237
	HIV1_M	80764.1	80829.5	80668.1	<0.001	<0.001	0.442
	HIV1_setC	180575.0	180702.3	180494.4	<0.001	<0.001	0.107
	HIV1_setD	298489.9	298695.3	298260.6	<0.001	<0.001	0.133
	HIV1_setE	289111.3	289292.1	288941.9	<0.001	<0.001	0.112
	HIV1_setF	214375.9	214692.2	214289.4	<0.001	<0.001	0.148
	EIV	126149	126365.4	125300	<0.001	<0.001	0.192
	BIV	24505.2	24506.9	24513.2	>0.05	>0.05	0.15
	FIV	164542.1	164487.9	164260.4	<0.001	<0.001	0.114
	CAV	351329.9	351871.5	350721.9	<0.001	<0.001	0.174
	SIV	110731.2	110816	110663.3	<0.001	<0.001	0.144
Filoviridae	Ebola_2	53147.3	53143.0	53149.9	>0.05	>0.50	0.264
Orthomyxo-viridae	Flu A 2	82872.8	83010.2	82849.7	<0.001	<0.001	0.27
Orthomyxo-viridae	Flu B 1	50090.4	50144.1	50060.9	<0.001	<0.001	0.311
Coronaviridae	SARS-COV1	214715.3	214968.5	214644.39	<0.001	<0.001	0.198
	SARS-COV2	15715.4.2	15715.6	15696.7	<0.001	<0.001	1.536
	SARB	573966.3	573815.1	572517.0	<0.001	<0.001	0.301
	MERS-COV	516683.2	516983.4	516608.9	<0.001	<0.001	0.169

Table 5

Relative rate change for C to A, G to A, A to T, G to C, T to G, and C to T mutations under the 11 degrees of non-reversibility alongside the maintained rates for A to C, A to G, T to A, C to G, G to T, and T to C.

Degree of non-reversibility (DNR)	Relative rates of different nucleotide substitution types (from-to)
Degree of non-reversibility (DNR)	C-A	A-C	G-A	A-G	A-T	T-A	G-C	C-G	T-G	G-T	C-T	T-C
0	0.166	0.166	1	1	0.14	0.14	0.131	0.131	0.118	0.118	1.101	1.101
2	2.166	0.166	3	1	2.14	0.14	2.131	0.131	2.118	0.118	3.101	1.101
4	4.166	0.166	5	1	4.14	0.14	4.131	0.131	4.118	0.118	5.101	1.101
6	6.166	0.166	7	1	6.14	0.14	6.131	0.131	6.118	0.118	7.101	1.101
8	8.166	0.166	9	1	8.14	0.14	8.131	0.131	8.118	0.118	9.101	1.101
10	10.166	0.166	11	1	10.14	0.14	10.131	0.131	10.118	0.118	11.101	1.101
12	12.166	0.166	13	1	12.14	0.14	12.131	0.131	12.118	0.118	13.101	1.101
14	14.166	0.166	15	1	14.14	0.14	14.131	0.131	14.118	0.118	15.101	1.101
16	16.166	0.166	17	1	16.14	0.14	16.131	0.131	16.118	0.118	17.101	1.101
18	18.166	0.166	19	1	18.14	0.14	18.131	0.131	18.118	0.118	19.101	1.101
20	20.166	0.166	21	1	20.14	0.14	20.131	0.131	20.118	0.118	21.101	1.101

Additional files

MDAR checklist: https://cdn.elifesciences.org/articles/87361/elife-87361-mdarchecklist1-v1.docx
Download elife-87361-mdarchecklist1-v1.docx
Supplementary file 1 Summary of the viral genome component datasets used in the study.: https://cdn.elifesciences.org/articles/87361/elife-87361-supp1-v1.docx
Download elife-87361-supp1-v1.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Article PDF

Open citations (links to open the citations from this article in various online reference manager services)

Mendeley

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

Rita Sianga-Mete
Penelope Hartnady
Wimbai Caroline Mandikumba
Kayleigh Rutherford
Christopher Brian Currin
Florence Phelanyane
Sabina Stefan
Steven Weaver
Sergei L Kosakovsky Pond
Darren P Martin

(2025)

Viral genome sequence datasets display pervasive evidence of strand-specific substitution biases that are best described using non-reversible nucleotide substitution models

eLife 12:RP87361.

https://doi.org/10.7554/eLife.87361.3

Share this article

Cite this article

Ternary plots illustrating the relative fit of the NREV12, NREV6, and GTR nucleotide substitution models based on weighted small sample corrected Akaike information criterion (AIC-c) scores for 30 dsDNA, 31 dsRNA, 33 ssDNA, and 47 ssRNA virus nucleotide sequence datasets.

Weighted Robinson-Foulds distances between inferred and true phylogenetic trees for datasets simulated with different degrees of nucleotide substitution non-reversibility and different average pairwise sequence identities (APIs) (~75%, ~80%, ~85%, ~90%, and ~95%).

Phylogenetic tree inferred from an alignment of real sequences (Avian Leukosis virus) that was used to simulate datasets with degrees of non-reversibility (DNRs) varying from 0 to 20.

Akaike information criterion (AIC) scores and likelihood ratio test (LRT) results for double-stranded DNA virus datasets.

Akaike information criterion (AIC) scores and likelihood ratio test (LRT) results for double-stranded RNA datasets.

Small sample corrected Akaike information criterion (AIC-c) scores and likelihood ratio test (LRT) results for single-stranded DNA datasets.

Small sample corrected Akaike information criterion (AIC-c) scores and likelihood ratio test (LRT) results for single-stranded RNA datasets.

Relative rate change for C to A, G to A, A to T, G to C, T to G, and C to T mutations under the 11 degrees of non-reversibility alongside the maintained rates for A to C, A to G, T to A, C to G, G to T, and T to C.

MDAR checklist

Supplementary file 1

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)