Tratopedia
繁中
Settings

Text size

Language

Theme

High contrast

Version

v1.230.2

The release this page was built from. It is what the service worker caches under.

Slide deck

Back

Six single-cell analysis platforms, compared

No single-cell analysis platform covers every kit and data type. 10x Cloud and Sentira take only 10x data; Trailmaker processes only Parse kits upstream; Partek Flow and BioTuring read many vendors’ formats; ROSALIND suits nCounter, GeoMx and bulk work.

  1. Research summary · 17 October 2026

    Six single-cell analysis platforms, compared

    Partek Flow · Trailmaker · 10x Cloud · 10x Sentira · BioTuring · ROSALIND

    Single-cell, spatial and bulk sequencing analysis without code1 / 15

  2. Basics

    Three kinds of data

    • scRNA-seq / snRNA-seq

      Single-cell sequencing

      • Which genes each cell expresses, measured cell by cell.
      • An analogy: taking a salad apart and checking every leaf.
    • Visium · Xenium · CosMx

      Spatial transcriptomics

      • Gene expression plus where each cell sits in the tissue section.
      • An analogy: knowing not only who came, but where they sat.
    • bulk RNA-seq

      Bulk sequencing

      • The whole sample measured at once: the average of all its cells.
      • An analogy: blending the salad into a juice and tasting it.

    Six single-cell analysis platforms · sources read 17 October 20262 / 15

  3. Basics

    Analysis runs upstream, then downstream

    Some platforms do one half; some do both.

    1. 1Raw dataFASTQ files from the sequencer
    2. 2Upstream: align and count10x Cloud; Trailmaker’s Pipeline module
    3. 3Count matrixEvery cell by every gene
    4. 4Downstream: interpretPartek Flow, BioTuring, ROSALIND, Sentira; Trailmaker’s Insights module

    An analogy: upstream turns a recording into a transcript; downstream reads the transcript and draws out the points.3 / 15

  4. Core conclusion

    No platform does everything

    First check whose kit made the data and what type it is, then choose the platform — a reading from the documentation that follows.

    Six single-cell analysis platforms · sources read 17 October 20264 / 15

  5. Six platforms

    The six platforms at a glance

    PlatformOwnerRole
    Partek FlowIlluminaThe widest data-type list, bulk to spatial
    TrailmakerParse Biosciences (QIAGEN)Free to academics; FASTQ to figures
    10x Cloud10x Genomics10x’s own cloud, for upstream processing
    10x Sentira10x GenomicsNew in 2026: GPU-accelerated spatial analysis
    BioTuringBioTuringThe widest imaging-spatial list; a large curated atlas
    ROSALINDROSALIND, Inc.Interpreting bulk and NanoString data

    Six single-cell analysis platforms · sources read 17 October 20265 / 15

  6. Platform 1 of 6 · Illumina

    Partek Flow

    Point-and-click analysis of bulk, single-cell, ATAC and spatial data

    • Strengths

      Partek Flow

      • The widest list of data types
      • Can be self-hosted: data need not leave the institution
      • Visium, Xenium and CosMx imports documented
    • Weaknesses

      Partek Flow

      • Single-cell/spatial and pathway tools sold as add-ons
      • No published price
      • A cloud successor, Connected Multiomics, since Dec 2025
      • Custom R/Python support not established
    • Illumina-hosted or self-hostedRuns
    • Subscription and add-ons, by quotePrice
    • Cores running many technologiesBest fit

    Six single-cell analysis platforms · sources read 17 October 20266 / 15

  7. Platform 2 of 6 · Parse Biosciences (QIAGEN)

    Trailmaker

    From FASTQ to figures, free to academics and Parse customers

    • Strengths

      Trailmaker

      • Free to academics and Parse customers
      • Guided quality control, easy to start
      • Imports 10x, BD, Seurat and AnnData data
      • Data stored in the EU (Ireland)
    • Weaknesses

      Trailmaker

      • The upstream pipeline takes Parse data only
      • No spatial or ATAC workflow documented
      • No in-platform scripting documented; objects can be downloaded
      • Aug 2026: a jury verdict against Parse; an appeal ruling against 10x patents
    • Cloud (AWS, Ireland)Runs
    • Free for academics and Parse customersPrice
    • Parse users, academic labsBest fit

    Six single-cell analysis platforms · sources read 17 October 20267 / 15

  8. Platform 3 of 6 · 10x Genomics

    10x Cloud

    10x’s own Cell Ranger and Space Ranger, run in the cloud

    • Strengths

      10x Cloud

      • The vendor’s own pipelines
      • 5 free analyses per dataset, 50 a month
      • Matrices and Loupe files stored free indefinitely
      • Sends data straight to BioTuring
    • Weaknesses

      10x Cloud

      • 10x data only
      • A processing service, not a full downstream workbench
      • Paid use only in the US and Canada
      • Spatial support is Visium HD
    • CloudRuns
    • Free quota, then from $5.00 an analysisPrice
    • Upstream processing for every 10x userBest fit

    Six single-cell analysis platforms · sources read 17 October 20268 / 15

  9. Platform 4 of 6 · 10x Genomics · announced 29 Sep 2026

    10x Sentira

    GPU-accelerated spatial analysis, paid from 2027

    • Strengths

      10x Sentira

      • GPU acceleration; claimed hours down to minutes
      • Free now for Xenium and Atera users
      • Built by the instrument’s maker
      • A single-cell module listed as coming soon
    • Weaknesses

      10x Sentira

      • Xenium and Atera data only, for now
      • Paid from early 2027; price not announced
      • Security and export details not public
      • Speed claim not independently verified
    • Cloud (GPU)Runs
    • Free in early access; credits from 2027Price
    • Xenium and Atera users, to trialBest fit

    Six single-cell analysis platforms · sources read 17 October 20269 / 15

  10. Platform 5 of 6 · BioTuring

    BioTuring

    The widest imaging-spatial list, and a 700M+ cell database queried in plain language

    • Strengths

      BioTuring

      • Xenium, CosMx, Visium HD and MERSCOPE named
      • Plain-language queries; reference-based cell typing
      • A notebook workbench, BioStudio, for those who code
      • 30-day trial for 10x customers new to it
    • Weaknesses

      BioTuring

      • No published price
      • The cloud version means uploading data to the vendor
      • Several product names to tell apart
      • Cell counts and speed are the vendor’s own figures
    • Cloud, with a workbench that reaches private serversRuns
    • By quotePrice
    • Large spatial datasets, pharma teamsBest fit

    Six single-cell analysis platforms · sources read 17 October 202610 / 15

  11. Platform 6 of 6 · ROSALIND, Inc.

    ROSALIND

    Interpretation and collaboration for bulk and NanoString data

    • Strengths

      ROSALIND

      • Published price: academic from $208 a month
      • Calls itself Bruker’s preferred nCounter solution
      • 50+ knowledge bases; strong pathway analysis
      • Collaboration spaces and project transfer
    • Weaknesses

      ROSALIND

      • No imaging single-cell spatial support listed
      • A non-binding term sheet over its platform assets, signed Mar 2026
      • Its AI direction has no release date
      • No independent benchmark found
    • CloudRuns
    • Published: academic $208–$416 a monthPrice
    • NanoString and bulk RNA-seq teamsBest fit

    Six single-cell analysis platforms · sources read 17 October 202611 / 15

  12. Side by side

    Which data can each one handle?

    PlatformSingle-cellATAC / multiomeSpatialBulkVendor-neutral
    Partek FlowYesYesVisium, Xenium, CosMxYesHigh
    TrailmakerYesNot documentedNot documentedNot documentedMedium (downstream)
    10x Cloud10x only10x multiomeVisium HDNo10x only
    10x SentiraComing soonNot statedXenium, AteraNot stated10x only
    BioTuringYesNot checkedThe widest listYesHigh
    ROSALINDYesBulk ATAC onlyGeoMx onlyA core useMedium (NanoString-leaning)

    Vendor-neutral: whether it can analyse data made with other vendors’ kits.12 / 15

  13. Side by side

    Deployment, AI and price

    PlatformDeploymentAI featuresPrice
    Partek FlowIllumina-hosted or self-hostedSuccessor has task suggestions and an AI methods summaryBy quote
    TrailmakerCloud (EU)Help chatbot, no data accessFree for academics
    10x CloudCloudPrompt-based interface; automated annotationFree quota
    10x SentiraCloud (GPU)Agentic single-cell module coming soonCredits from 2027
    BioTuringCloud, or a workbench reaching private serversPlain-language queries; reference cell typingBy quote
    ROSALINDCloudA stated direction onlyFrom $208 a month (academic)

    Where data are stored decides whether a platform meets an institution’s data-governance rules.13 / 15

  14. Choosing

    Pick by your situation

    Readings from the comparison, not rankings.

    Your situationSuggested platform
    Academic lab, 10x only, small budget10x Cloud with Loupe; Trailmaker for further analysis
    Parse kitsTrailmaker
    A core running many technologiesPartek Flow or Connected Multiomics
    Large imaging-spatial studies, pharmaBioTuring SpatialX and Talk2Data
    Already on Xenium or AteraTry Sentira free; keep a fallback
    Mostly NanoString or bulk dataROSALIND
    Bioinformaticians on the teamPair with Seurat or Scanpy; choose platforms that export

    Six single-cell analysis platforms · sources read 17 October 202614 / 15

  15. Before buying

    Three things to check first

    CheckWhy
    Few independent comparisonsSpeed and scale figures are the vendors’ own; trial two or three platforms on the same dataset.
    Not established herePartek custom scripts; BioTuring free-plan limits, prices and export formats; ROSALIND security certifications; where 10x Cloud stores data; GxP support. Ask the vendor in writing.
    Company eventsQIAGEN owns Parse (Dec 2025); the Delaware verdict and an injunction request (Aug 2026); the Rosalind term sheet (Mar 2026); Partek Flow’s cloud successor (Dec 2025); Sentira’s 2027 pricing.
    Checked and not carriedSentira as 10x’s first paid software (10x Cloud already bills); where 10x Cloud stores data; BioTuring’s free-plan limits; ROSALIND’s certifications and headcount.

    Next step: pick two or three platforms and run the same dataset through each.15 / 15