Reserve a spot

Every RNA-seq result, with the checks shown.

CodeISB runs your bulk RNA-seq analysis and audits it as it goes: swapped samples, the wrong strandedness setting, batch tangled with condition. You get the figures plus a report your reviewer can verify.

Reserve a founding spot See what the audit catches

No payment today. Founding labs get access first and keep their launch price for 12 months.

Audit report · run 4F2A916 samples · GRCh38
4 pass2 warn1 fail 7 checks shown
  • FAIL
    Strandedness
    Declared unstranded. Inferred reverse-stranded (96.9% of reads). Counts were redone with the correct setting.
  • WARN
    Batch and condition
    All 3 treated samples sit in batch 2. The treatment effect is only partly separable from batch.
  • WARN
    rRNA fraction
    Sample S5 at 14.2% against a cohort median of 3.1%.
  • PASS
    Sample identity
    6 of 6 samples match their declared sex-linked marker expression.
  • PASS
    Replicate correlation
    Lowest within-group Spearman 0.987.
  • PASS
    Library size
    21.4M to 28.9M reads, no outliers.
  • PASS
    Versions recorded
    Genome, annotation, DESeq2 and every parameter saved with the run.
Example report built from made-up data to show the format.
The problem

The mistakes that change a gene list are invisible in the figures.

01

Wrong strandedness

Library prep and counting settings disagree, and a large share of reads land on the wrong strand. The volcano plot still looks fine.

Check: inferred vs declared strand
02

Swapped or mislabeled samples

Two tubes trade names somewhere between the bench and the sample sheet. Differential expression runs without complaint.

Check: sex-linked and marker-gene identity
03

Batch tangled with condition

Every treated sample was sequenced on the same lane. The model fits, and the result may be the lane rather than the treatment.

Check: design matrix and batch overlap

A reviewer, a collaborator, or you in two years cannot see any of this from a figure. CodeISB runs the checks every time and puts the outcome next to the result.

How it works

Upload data, read the audit, publish with the record attached.

  1. Upload

    Your count matrix and sample sheet. Name the comparison you care about.

  2. Run

    Normalization, differential expression and pathway analysis run on a pinned, versioned pipeline.

  3. Audit

    Each check returns pass, warn or fail with the evidence. Fixable problems are fixed and logged.

  4. Export

    Figures, tables and the audit report, plus a run record that reproduces the analysis.

What you get

Results a reviewer can check without emailing you.

  • Publication-ready figuresPCA, sample distance, volcano, MA and heatmaps, as vector files.
  • Differential expression tablesPer-gene statistics with the model design and contrasts written out.
  • Pathway resultsGene set analysis with the gene set collection and version named.
  • The audit reportEvery check, its result and the evidence, as a PDF you can attach to a submission.
  • A run recordSoftware versions, reference files, parameters and a command that repeats the run.
  • Your data back, in open formatsCSV, TSV and standard objects. Nothing is locked in.
Founding pricing

Priced per project, so you can try it on one dataset.

Single project

$149per analysis

One dataset, one comparison set, one audit report. For a first paper or a one-off.

  • Full analysis and audit report
  • Figures and tables
  • Run record for reproduction
Reserve this plan
Most labs

Lab

$390per month

For a lab with regular sequencing. Up to 6 projects each month, shared across your team.

  • Everything in Single project
  • Up to 6 projects a month
  • Team workspace with saved settings
  • Audit history across projects
Reserve this plan

Core facility

$1,200per month

For a sequencing core or a group that delivers data to many labs.

  • Everything in Lab
  • Up to 30 projects a month
  • Branded audit reports for your clients
  • Priority support
Reserve this plan

Founding prices apply for 12 months from your first project. Reserving a spot does not charge you or commit you to buy.

Reserve a spot

Tell us how you run RNA-seq today.

We are opening access to a small group of founding labs first. Your answers decide what gets built first, so be specific.

  • No payment taken and no commitment
  • Founding price held for 12 months
  • We reply personally, usually within two working days
Plan you are leaning toward

We use your answers only to contact you about CodeISB access. We do not sell or share them.

Questions

Before you reserve.

Does this replace my bioinformatician?

No. It automates the routine analysis and the checks that people skip under deadline. A bioinformatician still decides the design, the contrasts and what the results mean.

What can I upload at launch?

A count matrix and a sample sheet. The form above asks what you start from so we can decide whether to add FASTQ processing first.

What if a check fails?

The report says what failed and shows the evidence. If the fix is mechanical, such as a strandedness setting, we apply it and log the change. If it needs a decision from you, the run stops and tells you why.

Which organisms are supported?

Human and mouse first. Tell us what you work on in the form and it will shape the order after that.

When can I start?

Founding labs are invited in small groups as the tool is ready. We will email you with a date and a free first run on your own data.