Amplicon sequence variant (ASV) count table for paired baseline (T0) and 6-month follow-up (T1) fecal samples from 15 patients with Myasthenia gravis (MG)

Chang, Che-Cheng 1 ; Lu, Chi-Jie2; Liu, Tzu-Chi2

Published Jul 06, 2026 on Dryad. https://doi.org/10.5061/dryad.hqbzkh1zg

Data files

Jul 06, 2026 version files 935.91 KB

asv_reps.fasta

548.21 KB
ASV_table_MG_T0_T1.xlsx

148.69 KB
asv_table.even.txt

233.46 KB
README.md

5.55 KB

Abstract

This dataset comprises bacterial 16S rRNA gene amplicon sequencing data generated to characterize the gut microbial community composition of human participants. Total genomic DNA was extracted from homogenized fecal material, and the V3–V4 hypervariable region was amplified using the 341F/806R primer pair and sequenced on an Illumina MiSeq platform (2 × 300 bp, paired-end). The deposit contains samples derived from 15 participants. Primary data are provided as demultiplexed raw paired-end reads in FASTQ format. Accompanying processed files—generated in QIIME 2 (v2021.4) with DADA2—include the amplicon sequence variant (ASV) feature table, representative ASV sequences, and taxonomic assignments made against the SILVA reference database. A sample metadata file links each sequence file to its corresponding sample.

These data support standard microbiome workflows, including α-diversity (observed ASVs, Shannon, Simpson) and β-diversity estimation, taxonomic profiling, and differential-abundance testing (e.g., LEfSe). Because the deposit provides exact, single-nucleotide-resolution ASVs rather than clustered OTUs, it is well suited to reuse in cross-study comparisons, meta-analyses, reference-classifier benchmarking, and methodological development. All fecal samples were obtained from human participants with written informed consent and approval by the Research Ethics Committee of Fu-Jen Catholic University Hospital (FJUH109054) and conducted in accordance with the Declaration of Helsinki. Sequence data and metadata have been de-identified, and no personally identifiable information is included.

Dataset DOI: 10.5061/dryad.hqbzkh1zg

Description of the data and file structure

Files and variables

File: asv_table.even.txt, ASV_table_MG_T0_T1.xlsx

Description: Same rarefied ASV data as the xlsx in long format (ASV × sample) with taxonomy, both time points combined. Tab-delimited UTF-8 text; 1,267 data rows (ASVs) × 32 columns (ASV ID + 30 samples + taxonomy). No missing values; 0 denotes a true zero count.

Variables:

#OTU_num: ASV identifier (matches ASV_* in the xlsx and the headers in asv_reps.fasta).
MG.02: baseline (T0) rarefied read count, patient 02; integer ≥ 0, unitless.
MG.05: baseline (T0) rarefied read count, patient 05; integer ≥ 0, unitless.
MG.06: baseline (T0) rarefied read count, patient 06; integer ≥ 0, unitless.
MG.07: baseline (T0) rarefied read count, patient 07; integer ≥ 0, unitless.
MG.09: baseline (T0) rarefied read count, patient 09; integer ≥ 0, unitless.
MG.10: baseline (T0) rarefied read count, patient 10; integer ≥ 0, unitless.
MG.11: baseline (T0) rarefied read count, patient 11; integer ≥ 0, unitless.
MG.12: baseline (T0) rarefied read count, patient 12; integer ≥ 0, unitless.
MG.13: baseline (T0) rarefied read count, patient 13; integer ≥ 0, unitless.
MG.14: baseline (T0) rarefied read count, patient 14; integer ≥ 0, unitless.
MG.15: baseline (T0) rarefied read count, patient 15; integer ≥ 0, unitless.
MG.16: baseline (T0) rarefied read count, patient 16; integer ≥ 0, unitless.
MG.17: baseline (T0) rarefied read count, patient 17; integer ≥ 0, unitless.
MG.22: baseline (T0) rarefied read count, patient 22; integer ≥ 0, unitless.
MG.23: baseline (T0) rarefied read count, patient 23; integer ≥ 0, unitless.
MGP.02: 6-month follow-up (T1) rarefied read count, patient 02; integer ≥ 0, unitless.
MGP.05: 6-month follow-up (T1) rarefied read count, patient 05; integer ≥ 0, unitless.
MGP.06: 6-month follow-up (T1) rarefied read count, patient 06; integer ≥ 0, unitless.
MGP.07: 6-month follow-up (T1) rarefied read count, patient 07; integer ≥ 0, unitless.
MGP.09: 6-month follow-up (T1) rarefied read count, patient 09; integer ≥ 0, unitless.
MGP.10: 6-month follow-up (T1) rarefied read count, patient 10; integer ≥ 0, unitless.
MGP.11: 6-month follow-up (T1) rarefied read count, patient 11; integer ≥ 0, unitless.
MGP.12: 6-month follow-up (T1) rarefied read count, patient 12; integer ≥ 0, unitless.
MGP.13: 6-month follow-up (T1) rarefied read count, patient 13; integer ≥ 0, unitless.
MGP.14: 6-month follow-up (T1) rarefied read count, patient 14; integer ≥ 0, unitless.
MGP.15: 6-month follow-up (T1) rarefied read count, patient 15; integer ≥ 0, unitless.
MGP.16: 6-month follow-up (T1) rarefied read count, patient 16; integer ≥ 0, unitless.
MGP.17: 6-month follow-up (T1) rarefied read count, patient 17; integer ≥ 0, unitless.
MGP.22: 6-month follow-up (T1) rarefied read count, patient 22; integer ≥ 0, unitless.
MGP.23: 6-month follow-up (T1) rarefied read count, patient 23; integer ≥ 0, unitless.
taxonomy: taxonomic lineage assigned against the SILVA database (QIIME 2 q2-feature-classifier); greengenes-style, semicolon-separated seven ranks k__; p__; c__; o__; f__; g__; s__ (kingdom, phylum, class, order, family, genus, species); unresolved ranks left empty after the prefix.

File: asv_reps.fasta

Description: FASTA file of representative nucleotide sequences for the ASVs (DADA2 in QIIME 2). Contains 1,287 records. Each record header is the ASV identifier (>ASV_n, matching #OTU_num / the ASV_* columns), followed by the representative 16S rRNA V3–V4 sequence. No variables (sequence file).

Code/software

Viewing the data files. No proprietary software is required. The tabular data can be opened in any spreadsheet application or read programmatically. The .xlsx workbook opens in any spreadsheet program; asv_table_even.txt is a UTF-8, tab-delimited text file readable in any text editor or imported into R (read.delim) or Python (pandas.read_csv(sep="\t")); asv_reps.fasta is a plain-text FASTA file viewable in any text editor or sequence viewer.

Software used to generate the data. Raw 16S rRNA V3–V4 reads were processed in QIIME 2 (v2021.4): primer/adapter trimming with the cutadapt plugin; denoising, dereplication, paired-end merging, and chimera removal with DADA2 (maxEE = 2), which produced the amplicon sequence variants (ASVs) and their representative sequences (asv_reps.fasta); and taxonomic classification with the q2-feature-classifier plugin trained against the SILVA reference database (taxonomy column). ASV count tables were rarefied to an even depth of 29,004 reads per sample to yield the files provided here.

Downstream analysis. Statistical analysis and modelling were performed in Python (v3.8.8) in Jupyter Notebook (v6.3.0)

Access information

N/A

Human subjects data

The study protocol was approved by the Research Ethics Committee of Fu-Jen Catholic University Hospital (FJUH109054) and conducted in accordance with the Declaration of Helsinki. All participants provided written informed consent.

Fecal sample collection and DNA extraction

Fecal samples were collected from each participant in sterile containers, snap-frozen within 2 h of voiding, and stored at –80 °C until processing. Total genomic DNA was isolated from approximately 200 mg of homogenized fecal material using the EasyPrep Stool Genomic DNA Kit (Biotools, New Taipei City, Taiwan), following the manufacturer's instructions. All extracts were normalized to 5 ng/μL before amplification.

16S rRNA gene library construction and sequencing

The V3–V4 hypervariable region of the bacterial 16S rRNA gene was amplified using the universal primer pair 341F and 806R, following the Illumina 16S Metagenomic Sequencing Library Preparation protocol. Each 25-μL reaction contained 12.5 ng of template DNA combined with KAPA HiFi HotStart ReadyMix (Roche, Basel, Switzerland). Amplicons (approximately 500 bp) were resolved on 1.5% agarose gels and purified using AMPure XP magnetic beads (Beckman Coulter, Brea, CA, USA). Dual-indexed sequencing adapters were appended in the second PCR step using the Nextera XT Index Kit (Illumina, San Diego, CA, USA). Library quality was monitored using a Qubit 4.0 fluorometer and Qsep100 capillary electrophoresis system (BiOptic, Taipei, Taiwan). The library was sequenced (2 × 300 bp paired-end) using an Illumina MiSeq platform.

16S rRNA sequencing analysis

Raw reads were demultiplexed using sample-specific barcodes, and primer and adapter sequences were trimmed using the QIIME 2 cutadapt plugin (v2021.4). Quality filtering, dereplication, parametric error-model learning, paired-end joining, and chimera removal were performed using DADA2 as implemented in QIIME 2, with a maximum expected error threshold of two per read (maxEE = 2) . DADA2 resolves single-nucleotide differences in merged amplicons, returning exact ASVs rather than clustered operational taxonomic units. Taxonomic assignments were performed using the q2-feature-classifier plugin trained against the SILVA database.

Community diversity and differential abundance analyses

ASV tables were rarefied to the minimum library depth across samples before diversity computation. Within-sample α-diversity was measured using the observed ASV count, Shannon's and Simpsons index. paired Wilcoxon signed-rank test. Differentially abundant taxa were identified using linear discriminant analysis effect size (LEfSe) . Subsequently, a linear discriminant analysis (LDA) is applied to the bacterial taxa identified as significantly different to determine the effect size of each differentially abundant taxon. In the present study, taxa with an LDA score > 2 were considered significant.