. 2022 Jun 30;25(7):104488. doi: 10.1016/j.isci.2022.104488

Table 5.

Weighted Gene Correlation Network Analysis (WGCNA) of genome-wide methylation of CpG sites that change significantly with initial HIV infection in the SC group

Co-methylation Module^a	# of CpGs in Module	Module eigenvector methylation^b			# (%) of CpGs with kME≥0.85^d
Co-methylation Module^a	# of CpGs in Module	Visit A Mean (SD)	Visit B Mean (SD)	p value^c	# (%) of CpGs with kME≥0.85^d
1	133,037	0.004 (0.017)	−0.006 (0.020)	3.6x10⁻⁴	37,284 (28)
2	37,328	−0.015 (0.014)	−0.003 (0.017)	5.5x10⁻⁸	8097 (21)
3	16,985	−0.008 (0.014)	0.003 (0.022)	5.5x10⁻⁸	2300 (14)
4	9,775	0.012 (0.018)	−0.005 (0.021)	5.8x10⁻⁹	1596 (16)
5	5,615	−0.022 (0.014)	0.004 (0.017)	3.2x10⁻²¹	672 (12)
6	5,184	0.009 (0.017)	−0.002 (0.024)	3.2x10⁻⁴	892 (17)
7	2,398	−0.006 (0.023)	0.005 (0.022)	2.2x10⁻⁴	514 (21)
8	1,677	0.016 (0.016)	−0.004 (0.019)	1.5x10⁻¹³	141 (8)
9	1,019	0.006 (0.019)	−0.005 (0.021)	3.7x10⁻⁵	50 (5)
10	962	0.016 (0.016)	−0.005 (0.020)	2.5x10⁻¹²	58 (6)
11	707	−0.013 (0.023)	−0.002 (0.022)	6.0x10⁻⁴	22 (3)
12	222	−0.012 (0.025)	0.003 (0.015)	1.8x10⁻⁶	1 (0.4)
13	215	−0.017 (0.027)	−0.005 (0.025)	2.0x10⁻⁴	26 (12)
14	106	−0.009 (0.014)	0.010 (0.022)	1.7x10⁻¹⁰	5 (4.7)
15	94	−0.005 (0.018)	0.015 (0.023)	1.1x10⁻¹⁰	1 (1)
16	67	−0.015 (0.018)	−0.003 (0.020)	2.8x10⁻⁶	1 (1.5)
17	44	−0.022 (0.021)	−0.002 (0.019)	5.9x10⁻¹¹	2 (5)
18	38	0.011 (0.007)	−0.003 (0.018)	2.5x10⁻¹¹	17 (45)

Each Co-methylation Module is a cluster of CpG methylation sites within the 850,00 + sites evaluated on the Infinium MethylationEPIC BeadChip, identified by WGCNA to be correlated with each other; any one CpG site belongs to only one Module. Out of a total of 67 Modules identified by WGCNA utilizing all samples from all participants at both visits (n = 407 samples), 18 Modules shown are those that are significantly associated with the change in HIV status from visit A to visit B in the SC group (initial HIV infection). Modules 1–18 are numbered according to the number of CpGs (largest to smallest) contained in each Module. All CpGs in Modules 1–18 are listed in Table S11 (Excel file).

Methylation levels are quantified by the beta value from the EPIC BeadChip assay, using the ratio of intensities between methylated and un-methylated alleles as described in the STAR Methods. In WGCNA, a representative methylation profile for each Module, known as the Module eigenvector, is defined as the first principal component in the Module methylation matrix. Mean and SD eigenvector methylation values shown for each Module are based on 102 HIV seroconverters (SC group) with observations at both Visits A and B.

Italicized p values are from a non-parametric group comparison test (Kruskal-Wallis) comparing mean Module eigenvector methylation from Visit A (before HIV infection) to Visit B (after initial HIV infection); level of significance for Module association with HIV infection accounting for multiple comparisons is p< 0.05/67 or<7.5x10⁻⁴.

kME is the intramodular connectivity measure for each CpG calculated from the WGCNA, and ≥0.85 is the threshold for a CpG to be considered a “hub” site, as described in the STAR Methods. All CpGs from Modules 1–18 with kME ≥0.85 were included in a pathways enrichment analysis (please see Table S12, Excel file), and Modules in bold (n = 5) contain at least one CpG in a gene that falls within biological pathways with significant p values after adjustment for multiple comparisons.