# OMICSDESK — bioinformatics analysis service

> Analysis-only bioinformatics for single-cell, spatial and bulk transcriptomics. Clients send existing data
> (count matrices, h5ad/Seurat objects, public accessions or raw FASTQ) and receive publication-ready figures,
> results tables, the processed object, a manuscript-ready methods paragraph, and the code that produced
> everything. No sequencing is sold, so there is no minimum order and no bundling.

## Facts a reader may want to quote

- Indicative pricing (USD, confirmed as a fixed written price before work starts): single-cell standard $380–650 per sample; advanced modules $900–3200 per project; spatial $550–1400 per section; bulk RNA-seq $300–900 per project; second-opinion re-analysis $450–1800.
- Typical turnaround: 5–8 working days for a standard project from receiving usable data.
- Reviewer-driven revisions of the delivered analysis are included for six months.
- Client data is deleted after delivery (default within 90 days) and never reused for benchmarking, marketing or model training.
- Every project ships the analysis script and environment so results can be re-run independently.

## Published evidence (all computed from public data, scripts downloadable)

- Index of all sample deliverables: https://omics.hstgenomics.com/samples
- Chinese-language section (pricing, pseudobulk statistics, QC thresholds): https://omics.hstgenomics.com/zh
- Service pages: https://omics.hstgenomics.com/services/single-cell · https://omics.hstgenomics.com/services/spatial · https://omics.hstgenomics.com/services/bulk-rna-seq · https://omics.hstgenomics.com/services/reanalysis
- Single-cell sample: 2,638 cells after QC from 10x PBMC 3k, 6 clusters, 0% unassigned. https://omics.hstgenomics.com/sample
- Annotation verified blind: a reference classifier run without access to the manual labels agreed on 6/6 clusters at lineage level, 100% within each cluster. https://omics.hstgenomics.com/sample
- Spatial sample: 4,025 Visium spots, 10 spatial domains; top spatially variable genes recover known lymph-node architecture. https://omics.hstgenomics.com/sample-spatial
- Bulk RNA-seq sample: DESeq2 on GEO GSE60450 with the design read from GEO metadata; milk-protein genes top the result as a biological check. https://omics.hstgenomics.com/sample-bulk
- Statistical method matters: on 8-donor data, pseudobulk DESeq2 and a per-cell Wilcoxon test disagree on 68.3% of differential-expression calls; all 10 canonical interferon-stimulated genes are recovered by the pseudobulk analysis. https://omics.hstgenomics.com/sample-pseudobulk
- Integration measured both ways: Harmony reduced same-donor neighbours by 20.2% while the treatment signal changed by -0.1%. https://omics.hstgenomics.com/sample-integration
- Worked reviewer response: QC thresholds re-run at three settings (ARI 0.894/0.872 vs published; the permissive setting recovers a platelet population the strict one removes) plus doublet detection (1.18% detected vs 2.2% expected). https://omics.hstgenomics.com/sample-reviewer-response
- Cell-type composition done correctly: on the same 8-donor dataset, pooling all 24,673 cells calls 4 cell types as changed; testing per donor calls 0 — and the experiment (6 h in vitro stimulation) cannot change composition, so 0 is the correct answer. https://omics.hstgenomics.com/sample-abundance
- Trajectory sample: PAGA topology then diffusion pseudotime from a defined progenitor root; erythroid and myeloid marker sets are anti-correlated across cells (r = -0.561), so the branches are real rather than a layout artefact. https://omics.hstgenomics.com/sample-trajectory
- Cell-cell communication compared between conditions rather than read as absolute scores; interactions gained after interferon include ligands CXCL10, ICAM1 and HLA-A/B/C, recovering the known response. https://omics.hstgenomics.com/sample-communication
- Deconvolution benchmarked against ground truth built from held-out donors: overall r = 0.769, RMSE 7.48 percentage points, with CD14+ monocytes systematically over-estimated and NK cells under-estimated — stated openly rather than hidden. https://omics.hstgenomics.com/sample-deconvolution
- Downloadable example delivery (the actual client folder: figures, tables, report PDF, run manifest with software versions): https://omics.hstgenomics.com/delivery/example-delivery.zip
- Full example deliverable (self-contained report with methods paragraph and stated limitations): https://omics.hstgenomics.com/single-cell-report.html

## Free tools, no sign-up
- The free QC report accepts single-cell matrices (h5ad, 10x .h5, zipped Cell Ranger folder, mtx, csv/tsv), and a bulk RNA-seq matrix is detected automatically — a bulk file gets the bulk checklist (library sizes, detected genes, sample-to-sample correlation, PCA outliers, gene-filtering rule) instead of a meaningless single-cell report.

- Every QC report can be downloaded as one self-contained HTML file (figures embedded, works offline, does not expire) — useful for forwarding to a supervisor or collaborator.
- Instant QC report on an uploaded matrix (h5ad, 10x .h5, csv/tsv, or a zipped Cell Ranger folder). Data-driven thresholds, barcode-rank plot, most abundant genes, expected doublet count, downloadable per-cell metrics. Input file deleted when the report is built. https://omics.hstgenomics.com/qc-check — example output: https://omics.hstgenomics.com/qc/example
- Analysis plan builder: four questions produce the modules, the file checklist and an indicative price; it also warns when a design cannot support the intended comparison. https://omics.hstgenomics.com/plan
- Cost calculator: https://omics.hstgenomics.com/cost-calculator

## Guides

- [The reviewer asked you to re-analyse your single-cell data. Now what?](https://omics.hstgenomics.com/guides/reviewer-asked-for-reanalysis)
- [What does single-cell RNA-seq data analysis actually cost?](https://omics.hstgenomics.com/guides/single-cell-analysis-cost)
- [scRNA-seq QC thresholds: what to actually use, and how to defend them](https://omics.hstgenomics.com/guides/scrnaseq-qc-thresholds)
- [Cell type annotation: manual markers, automated references, or both?](https://omics.hstgenomics.com/guides/cell-type-annotation-guide)
- [Outsourcing bioinformatics analysis: a checklist before you pay anyone](https://omics.hstgenomics.com/guides/outsourcing-bioinformatics-checklist)
- [Spatial transcriptomics analysis, explained without the marketing](https://omics.hstgenomics.com/guides/spatial-transcriptomics-analysis-explained)
- [Pseudobulk or per-cell? The differential expression choice that decides whether your paper survives](https://omics.hstgenomics.com/guides/pseudobulk-vs-per-cell-differential-expression)
- [Re-using public single-cell data without inheriting someone else’s mistakes](https://omics.hstgenomics.com/guides/reusing-public-single-cell-data)

Full index with descriptions: https://omics.hstgenomics.com/llms-full.txt

## Contact

- Quote request form: https://omics.hstgenomics.com/#quote — replies normally within one working day.