Page 120 - Read Online

Wang et al. Microbiome Res Rep. 2025;4:23 https://dx.doi.org/10.20517/mrr.2024.94 Page 5 of 19

Library preparation
Twenty microliters of fecal sample DNA were mixed evenly with 2.5 µL of DNA from Pseudoalteromonas
sp. APC 3896 (4.31 ng/µL) and 2.5 µL of DNA from Planococcus sp. APC 3900 (9.59 ng/µL). The DNA
concentration was quantified using Qubit 4 and normalized to 5 ng/μL with 10 mm Tris before library
6
6
preparation. In this study, 2.32 × 10 copies of Pseudoalteromonas sp. APC 3896 and 5.99 × 10 copies of
Planococcus sp. APC 3900 were added to achieve a final relative abundance of the spike-in between 0.1%-
10% in most samples. The DNA samples were amplified, targeting the V3 and V4 regions of the 16S rRNA
gene using primers 341F/806R. This process produced an amplification fragment of approximately 465 bp.
Following the manufacturer's instructions, the amplicons were prepared for sequencing and analyzed using
the Illumina HiSeq 2000 platform with 2 × 250 bp chemistry (Illumina Technologies, USA).

Computational analysis
The raw paired-end reads were analyzed using the dada2 pipeline. Denoising and pre-processing of the
sequence reads were performed using the dada2 pipeline , while filtered and trimmed sequence data were

[30]
processed using the core sample inference algorithm. Taxonomy assignment to the Amplicon Sequence
Variants (ASVs) was conducted using the SILVA database release 138.1 . Diversity analysis was performed
[31]
in R using phyloseq and microbiome package.

Statistical analysis
SPSS 25.0 (Stanford, CA, USA) was used to perform statistical testing. The data were all tested for normal
distribution before comparison . For comparing the two groups of data, the Shapiro-Wilk test was used to
[32]
[33]
assess normality . If the data followed a normal distribution (P > 0.05), an independent sample t-test was
employed. If the data did not conform to a normal distribution (P < 0.05), the non-parametric Mann-
Whitney U test was used . For comparisons involving three or more groups, normality and homogeneity
[34]
of variances were first tested. If these assumptions were met, one-way ANOVA followed by Post hoc
Tukey’s test was used. If the data did not meet normality or homogeneity of variance assumptions, the non-
parametric Kruskal-Wallis test was applied . For correlation analysis, the Shapiro-Wilk test was used to
[35]
assess normal distribution. If the data were normally distributed, the Pearson correlation coefficient was
[36]
used; otherwise, the Spearman correlation coefficient was used .
RESULTS
Given the increasing need for accurate quantification of microbial abundances in microbiome studies,
particularly in relation to 16S rRNA and metagenomic sequencing outputs, we evaluated various
approaches to establish a reliable method for absolute quantitation. Traditional relative abundance
measures, while useful, often fail to capture the true variability in microbial load, leading to potential
misinterpretations of microbiome dynamics. To address this limitation, we implemented a DNA spike-in
quantification method that enables the determination of absolute microbial counts, providing a more
comprehensive view of microbial shifts and interactions. In this study, we compared the performance of our
spike-in method against conventional absolute quantification techniques such as flow cytometry, total DNA
quantification, qPCR, and plate count assessment. This approach allowed us to explore differences in
microbial loads between mothers and infants and examine how absolute quantification can alter
microbiome analysis outcomes, thereby emphasizing the necessity for absolute measures in microbiome
research.

Experimental design
Figure 1 provides an overview of the DNA spike-in quantification method and its comparison with other
alternative approaches. The initial step in the DNA spike-in procedure involved the extraction of DNA
from the spike-in bacteria Pseudoalteromonas sp. APC 3896 and Planococcus sp. APC 3900, as well as from