Processed 16S rRNA gene amplicon data (ASV table, taxonomy and representative sequences) for desert rock varnish, Al-Jafer Basin, Ma'an, Jordan

This dataset contains processed 16S rRNA gene amplicon data describing the prokaryotic communities of six desert rock-varnish libraries collected at three sampling points in the semi-arid Al-Jafer Basin near Ma'an, southern Jordan. Raw paired-end reads are available under NCBI BioProject PRJNA1526451. The deposited files comprise ASV abundance tables, SILVA v138.2 taxonomic assignments, representative ASV sequences, read-quality and read-retention summaries, diversity and richness tables, rarefaction curves, per-cycle quality profiles, and the six scripts that document the sequence-processing workflow. Primers were removed with cutadapt and reads were processed with DADA2 1.38.0 by paired-end denoising, merging and chimera removal, followed by taxonomic assignment against the DADA2-formatted SILVA v138.2 training set. The dataset is predominantly bacterial but also contains three archaeal ASVs (Nitrososphaeraceae) carrying 0.105% of retained reads; the 515F-926R primer pair amplifies both domains. Files File Content asv_abundance_counts.tsv read counts per ASV (rows) and library (columns), chimera-free asv_abundance_relative_percent.tsv the same table as a percentage of each library's reads asv_abundance_counts_min0.1pct.tsv counts after setting to zero any ASV below 0.1% relative abundance in each library; ASVs left at zero in all six libraries were dropped from the table, so it holds fewer rows than the unfiltered one asv_taxonomy.tsv SILVA v138.2 assignment per ASV, kingdom to genus (minBoot 80) asv_representative_sequences.fasta the merged ASV sequence for each identifier table4_diversity.csv per-library ASV richness, Shannon entropy, Simpson D, effective numbers of order 1 and 2, and evenness (Table 4 of the data article) table5_richness.csv distinct taxa per library at each rank and the percentage of reads with no genus-level assignment (Table 5 of the data article) rarefaction_curves.csv expected ASV richness per library at each subsampling depth, including each library's own retained-read total as its final depth rarefaction_asv.png the rarefaction curves plotted; the filled marker is each library at its full sequencing depth (Figure 1 of the data article) track_reads.csv reads surviving each processing stage, per library read_quality.csv Q20, Q30 and 5′-end primer-match percentages measured from the deposited FASTQ files quality_profiles_per_cycle.png mean quality score per sequencing cycle for both mates of the six libraries; these profiles were used together with the expected paired-end overlap requirement to inform the DADA2 truncation parameters. Deposited here only: the data article carries a single numbered figure, the rarefaction curves 00_test_truncation_params.R the probe run that compared three truncation and maxEE settings on the lowest-quality library 01_reads_and_quality.py retrieval of the six libraries from the archive and measurement of read quality 02_cutadapt.sh primer removal 03_run_dada2.R filtering, denoising, merging and chimera removal 04_assign_taxonomy.R taxonomic assignment and export of the tables above 05_diversity_and_richness.py diversity, effective numbers, rarefaction and per-rank richness How the tables were produced Primers were removed with cutadapt 5.2 (anchored 515F on R1, 926R on R2, `--discard-untrimmed`, e = 0.15). Reads were then processed with DADA2 1.38.0: `filterAndTrim(truncLen = c(230, 190), maxEE = c(3, 4), truncQ = 2, maxN = 0, rm.phix = TRUE)`, error models learned separately for the two flowcells (HCCKMBCX3 and HCMY3BCX3), `dada`, `mergePairs` (minOverlap 12, no mismatch), and `removeBimeraDenovo(method = "consensus")`. The truncation lengths were selected to balance declining 3′-end read quality against preservation of sufficient paired-end overlap for read merging; the probe run in `00_test_truncation_params.R` and the profiles in `quality_profiles_per_cycle.png` record the comparison. Taxonomy was assigned with `assignTaxonomy(minBoot = 80, tryRC = TRUE)` against the DADA2-formatted SILVA v138.2 training set, downloaded as `silva_nr99_v138.2_toGenus_trainset.fa.gz` (md5 1764e2a36b4500ccb1c7d5261948a414). The file was renamed locally to `refs/silva_toGenus.fa.gz`, which is the path `04_assign_taxonomy.R` reads; rename it the same way or edit that path. 356 ASVs and 134,647 reads remain after chimera removal. The 0.1% per-library cut-off retains 207 ASVs and 97% of the reads. Composition Of the 356 ASVs, 352 are assigned to Bacteria and 3 to Archaea (Nitrososphaeraceae), with 1 unassigned at kingdom level. The dataset is predominantly bacterial: archaeal ASVs carry 0.105% of retained reads. The 515F-926R primer pair amplifies both domains, so archaeal reads are expected. Identifiers ASV identifiers (ASV001 onwards) are local to this deposit, are ordered by the sequence table, and are stable across the abundance, taxonomy and sequence files. Provenance and credit The raw paired-end reads these tables derive from are deposited in the NCBI Sequence Read Archive under BioProject PRJNA1526451, which records the full author list of the associated data article. The files in this record are the processed tables produced from those reads during revision of that article and should be cited together with it.