Dataset DOI: 10.5061/dryad.2z34tmq0z

Description of the data and file structure

Data from: Interspecies Interactions Drive Community-Level Selection in Microbial Coalescence

Files and variables

• coalescence_data.zip — 1.93 GB (primary data archive containing raw FASTQ files)
• Sample_Sheet.xlsx — consolidated sample sheet (Metadata for data in coalescence_data.zip )
• Supplementary_Data_Tables.xlsx — 1.92 MB (consolidated processed-data workbook: ASV relative-abundance tables, per-community diversity measures, per-coalescence-event compositional metrics, and simulated coalescence event tables)
• README.md — 2.02 KB (original Dryad README)

Coalescence_data( coalescence_data.zip)

The archive contains:

• Raw sequencing data in FASTQ format (paired-end Illumina reads, one pair of files per sample).

FASTQ naming convention

FASTQ files follow the pattern:

<SampleID>_L<lane number>_R<read direction>_001.fastq.gz

Where:

• SampleID = unique sample identifier — matches the sample_id column in the sample sheet (e.g. P1-25)
• L = lane number (e.g., L001)
• R1 = forward read, R2 = reverse read
• 001 = file part index (often 001)

Example: P1-25_L001_R1_001.fastq.gz and P1-25_L001_R2_001.fastq.gz

Sample sheet (Sample_Sheet.xlsx)

A three-sheet Excel workbook provided as a top-level file alongside coalescence_data.zip. Contains per-sample metadata, coalescence pairings, and ASV taxonomy. The sample_id column is the primary key linking this file to the FASTQ filename prefixes inside coalescence_data.zip.

Sheet 1: samples — 588 rows, one per sequenced community

Per-sample metadata for all sequenced communities, including experimental design, coalescence parentage, and endpoint phenotypic measurements.

• sample_id = unique sample identifier (e.g., P1-25); matches the FASTQ filename prefix
• community_origin = Synthetic (12-species defined pool) or Natural (environmental community)
• medium = nitrogen level: Nutr- (low N), Base (medium N), or Nutr+ (high N)
• sample_type = Subcommunity (single-ancestor) or Coalescence (mixture of two subcommunities)
• replicate = biological replicate index (1 or 2)
• community_idx = community number within its (origin, medium, sample_type, replicate) group
• timepoint = Final (endpoint of the assembly experiment; only timepoint included)
• parent_1_community_idx, parent_2_community_idx = for Coalescence rows only, the community_idx of the two parent subcommunities (resolved via coalescence_recipe); empty for Subcommunity rows
• parent_1_sample_id, parent_2_sample_id = for Coalescence rows only, the sample_id of the two parent subcommunities (same medium and replicate)
• OD_final_mean, OD_final_std = mean and standard deviation of seven OD600 endpoint readings (fieldOD1…fieldOD7)
• pH_final_mean, pH_final_std = mean and standard deviation of seven endpoint pH readings (fieldPH1…fieldPH7)
• growth_curve_AUC_mean = mean of three growth-curve area-under-curve measurements (fieldGC1…fieldGC3)
• fieldOD1…fieldOD7 = raw endpoint OD600 readings (technical replicates)
• fieldPH1…fieldPH7 = raw endpoint pH readings (technical replicates)
• fieldGC1…fieldGC3 = raw growth-curve AUC measurements (technical replicates)
• notes = reason for data-quality flag (empty if none); possible values: no reads in coalescence result, no file, MN E7 missing, MN24 CC, excess unwanted ASVs (likely mislabel), ASV not in subcommunities at >0.3 abundance. Flagged samples are excluded from downstream analyses in the accompanying paper.

Sheet 2: coalescence_recipe — 62 rows

Mapping of coalescence events to their parent subcommunities. Each row defines one mixing event.

• event_id = sequential index across both origins (1–62)
• community_origin = Synthetic (47 events) or Natural (15 events)
• coalescence_community_idx = the community_idx of the resulting coalescence community
• parent_1_community_idx, parent_2_community_idx = the community_idx values of the two parent subcommunities that were mixed; resolved to specific sample_ids in the samples sheet via the (origin, medium, replicate, community_idx) tuple

Sheet 3: asv_taxonomy — 167 rows (43 synthetic + 124 natural)

Taxonomic assignments and representative sequences for all ASVs detected across the two experiments.

• community_origin = Synthetic or Natural; indicates which experiment this ASV is part of
• asv_id = ASV label (e.g., ASV1); numbering is independent between origins (synthetic ASV1 ≠ natural ASV1)
• kingdom, phylum, class, order, family, genus = taxonomic assignment from GreenGenes (16S rRNA gene)
• unique_sequence = representative 16S V4 amplicon nucleotide sequence for the ASV (~253 bp)

File: Supplementary_Data_Tables.xlsx

Description: Ten-sheet Excel workbook, preceded by a README index sheet, containing the processed data tables underlying the figures in the manuscript. It complements the raw sequencing reads in coalescence_data.zip and the sample-level metadata in Sample_Sheet.xlsx.

• ASV_Abundance_Synthetic (461 rows) Per-sample ASV relative abundances for the synthetic experiment.
• ASV_Taxonomy_Synthetic (43 rows) Taxonomy and representative sequence for each synthetic ASV.
• ASV_Abundance_Natural (126 rows) Per-sample ASV relative abundances for the natural experiment.
• ASV_Taxonomy_Natural (124 rows) Taxonomy and representative sequence for each natural ASV.
• Communities_Synthetic (462 rows) Per-community metadata, OD, pH, AUC, and diversity metrics for the synthetic experiment.
• Communities_Natural (126 rows) Per-community metadata, OD, pH, AUC, and diversity metrics for the natural experiment.
• CoalescenceEvents_Synthetic (282 rows) Per-event metadata, OD, pH, and coalescence metrics for the synthetic experiment.
• CoalescenceEvents_Natural (90 rows) Per-event metadata, OD, pH, and coalescence metrics for the natural experiment.
• SimulatedEvents (900 rows) Simulated generalized Lotka–Volterra (gLV) coalescence events from the primary model.
• SimulatedEvents_Uniform (900 rows) Simulated coalescence events from the uniform-interaction sensitivity variant; cosine metrics are omitted

Access information

The files will be available for public download after the dataset has been approved and published by Dryad.