Skip to content

🚀 Quick Start: Run with Docker

This tutorial walks through the cartloader workflow — all packaged inside a prebuilt Docker image with dependencies and input data included.

Use Cases

This is the fastest and simplest way to try cartloader — no setup, installation, or data download required.

Requirements

Users will need to:

Set Up Docker

If you are new to Docker, please refer to the Docker documentation for installation and basic usage.

Verify whether Docker is properly set up on your system: 

1
2
3
4
5
# Check if Docker is installed and show its version
docker --version

# Test if Docker can successfully run a container
docker run hello-world

If these commands fail, install Docker in your system.

Input Data

The input is an mouse hippocampus SGE in a FICTURE-compatible format compatible, prepared by sge_convert in cartloader .

File Format

transcripts.unsorted.tsv.gz: transcript-indexed SGE in TSV 
1
2
3
4
X        Y        gene     count
295.29   1422.35  Myo3a    0
1745.54  1110.72  Med14    1
1745.54  1110.72  Ntpcr    1
  • X: X coordinates in um
  • Y: Y coordinates in um
  • gene: gene symbols
  • count: expression count per pixel per gene
feature.clean.tsv.gz: UMI counts on a per-gene basis in TSV 

1
2
3
4
gene           gene_id             count
Gm29155        ENSMUSG00000100764  1
Pcmtd1         ENSMUSG00000051285  431
Gm26901        ENSMUSG00000097797  1
* gene: gene symbols * gene_id: gene IDs * count: expression count per gene

coordinate_minmax.tsv: X Y min/max coordinates 
1
2
3
4
xmin    0.14
xmax    2359.90
ymin    0.23
ymax    1439.95
  • xmin xmax: min and max X coordinates in um
  • ymin ymax: min and max Y coordinates in um

Data Access

Input data already included

The example dataset is preloaded in the Docker image — no need to download separately.

If needs, it is also available on Zenodo: DOI: 10.5281/zenodo.15701393

Set Up the Environment

Fixed paths in the Docker Image

Tools and dependencies have fixed paths in the Docker image (e.g., /usr/local/bin/pmtiles), which are used directly in the commands below. Skip specifying them manually.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
# ====
# Replace each placeholder with the actual path on your system.  
# ====

work_dir=/path/to/work/directory        # path to work directory that contains the downloaded input data
cd $work_dir

# Number of jobs
n_jobs=10                               # If not specify, the number of jobs defaults to 1.

# Unique identifier for your dataset
DATA_ID="seqscope_hippo"                # change this to reflect your dataset name
PLATFORM="seqscope"                     # platform information

# LDA parameters
train_width=18                           # define LDA training hexagon width (comma-separated if multiple widths are applied)
n_factor=6,12                            # define number of factors in LDA training (comma-separated if multiple n-factor are applied)

FICTURE Analysis

Compute spatial factors using punkst (FICTURE2 mode). See more details in Reference page.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
docker run -it --rm \
  -v $(pwd):/data \
  weiqiuc/cartloader:20250708b \
  run_ficture2 \
    --makefn run_ficture2.mk \
    --main \
    --in-transcript /app/data/sge/transcripts.unsorted.tsv.gz \
    --in-feature /app/data/sge/feature.clean.tsv.gz \
    --in-minmax /app/data/sge/coordinate_minmax.tsv \
    --cmap-file /app/cartloader/assets/fixed_color_map_256.tsv \
    --exclude-feature-regex '^(mt-.*$|Gm\d+$)' \
    --out-dir /data/ficture2 \
    --width ${train_width} \
    --n-factor ${n_factor} \
    --ficture2 /app/cartloader/submodules/punkst \
    --spatula /app/cartloader/submodules/spatula/bin/spatula \
    --n-jobs ${n_jobs} \
    --threads ${n_jobs}
Parameter Required Type Description
--main required 1 flag Enable cartloader to run all five steps
--in-transcript required string Path to input transcript-level SGE file
--out-dir required string Path to output directory
--width required int or comma-separated list LDA training hexagon width(s)
--n-factor required int or comma-separated list Number of LDA factors
--makefn string File name for the generated Makefile (default: run_ficture2.mk )
--in-feature string Path to input feature file
--in-minmax string Path to input coordinate min/max file
--cmap-file string Path to color map file
--exclude-feature-regex regex Pattern to exclude features
--spatula string Path to the spatula binary (default: spatula)
--ficture2 string Path to the punkst directory (defaults to punkst repository within submodules directory of cartloader)
--n-jobs int Number of parallel jobs (default: 1)
--threads int Number of threads per job (default: 1)

1: cartloader requires the user to specify at least one action. Available actions includes: --tile to run tiling step; --segment to run segmentation step; --init-lda to run LDA training step; --decode to run decoding step; --summary to run summarization step; --main to run all above five actions.

cartloader Compilation

Generate pmtiles and web-compatible tile directories. See more details in Reference page.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
docker run -it --rm \
  -v $(pwd):/data \
  weiqiuc/cartloader:20250708b \
  run_cartload2 \
    --makefn run_cartload2.mk \
    --fic-dir /data/ficture2 \
    --out-dir /data/cartload2 \
    --id ${DATA_ID} \
    --spatula /app/cartloader/submodules/spatula/bin/spatula \
    --pmtiles /usr/local/bin/pmtiles \
    --tippecanoe /usr/local/bin/tippecanoe \
    --n-jobs ${n_jobs} \
    --threads ${n_jobs}
Parameter Required Type Description
--fic-dir required string Path to the input directory containing FICTURE2 output
--out-dir required string Path to the output directory for PMTiles and web tiles
--id required string Dataset ID used for naming outputs and metadata
--makefn string File name for the generated Makefile (default: run_cartload2.mk)
--spatula string Path to the spatula binary (default: spatula)
--pmtiles string Path to the pmtiles binary (default: pmtiles)
--tippecanoe string Path to the tippecanoe binary (default: tippecanoe)
--n-jobs int Number of parallel jobs (default: 1)
--threads int Number of threads per job (default: 1)

Upload to Data Repository

Choose a data repository to host/share your output

cartloader supports two upload options (AWS and Zenodo) for storing PMTiles of SGE and spatial factors in a data repository.

Choose the one that best suits your needs.

AWS Uploads

Upload the generated cartloader outputs to your designated AWS S3 directory:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
# AWS S3 target location for cartostore
AWS_BUCKET="EXAMPLE_AWS_BUCKET"         # replace EXAMPLE_AWS_BUCKET with your actual S3 bucket name

docker run -it --rm \
  -v $(pwd):/data \
  weiqiuc/cartloader:20250708b \
  upload_aws \
    --in-dir /data/cartload2 \
    --s3-dir "s3://${AWS_BUCKET}/${DATA_ID}" \
    --aws /usr/local/bin/aws \
    --n-jobs ${n_jobs}
Parameter Required Type Description
--in-dir required string Path to the input directory containing the cartloader compilation output
--s3-dir required string Path to the target S3 directory for uploading
--aws string Path to the AWS CLI binary
--n-jobs int Number of parallel jobs

Zenodo Uploads

Upload the generated cartloader outputs to your designated Zenodo deposition or a new deposition.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
zenodo_token=/path/to/zenodo/token/file    # replace /path/to/zenodo/token/file by path to your zenodo token file

docker run -it --rm \
  -v $(pwd):/data \
  weiqiuc/cartloader:20250708b \
  upload_zenodo \
    --in-dir /data/cartload2 \
    --upload-method catalog \
    --zenodo-token ${zenodo_token} \
    --create-new-deposition \
    --title  "Yur Title" \
    --creators "Your Name" \
    --description "This is an example description"
Parameter Required Type Description
--in-dir required string Path to the input directory containing the cartloader compilation output
--upload-method required string Method to determine which files to upload. Options: all to upload all files in --in-dir; catalog to upload files listed in a catalog YAML file, user_list to upload files explicitly listed via --in-list
--catalog-yaml string Required if --upload-method catalog. Path to the catalog.yaml file generated in run_cartload2. If absent, will use the catalog.yaml in the input directory specified by --in-dir.
--zenodo-token required string Path to your Zenodo access file
--create-new-deposition flag a new Zenodo deposition will be created.
--title required string Required if --create-new-deposition. Title for the new Zenodo deposition.
--creators required list of str List of creators in "Lastname, Firstname" format.

Output Data

See more details of output at the Reference pages for run_ficture2 and run_cartload2.

Spatial Factor Inference from FICTURE

Below is an example of spatial factor inference results from FICTURE using a training width of 18, 12 factors, a fit width of 18, and an anchor resolution of 6.

FICTURE cmap

Factor RGB Weight PostUMI TopGene_pval TopGene_fc TopGene_weight
0 255,101,101 0.57145 2577934 Snap25,Cck,Hpca,Syt1,Atp1b1,Selenow,Ywhah,Scg5,Cnih2,Atp1a3,Vsnl1,Gnas,Cpe,Prkcb,Stmn2,Atp2b1,Gng3,Prkcg,Snrpn,Norad Nr2c1,Calca,Wdr91,Ccbe1,Esyt1,Epha3,Slc9a4,Thap7,Mapk11,Klk8,Arhgef6,1110018N20Rik,Iba57,Klk10,Gjd2,Leng9,Zbtb46,Slc9a2,Akr1c18,Dlx6os1 Snap25,Tmsb4x,Atp1b1,Actb,Cpe,Ywhah,Nrgn,Ppp3ca,Selenow,Calm1,Atp1a3,Calm2,Fth1,Rtn1,Cox8a,Gnas,Camk2a,Norad,Aldoa_v1,Ndrg4
1 237,238,0 0.15458 697338 Ppp3ca,Nrgn,Ptk2b,Olfm1,Ppp3r1,Gria2,Ncdn,Nsf,Syne1,Snca,Chn1,Tmsb4x,Rasgrf1,Grin2a,Enc1,Kalrn,Wasf1,Camk2b,Calm2,Epha7 Vwa8,Zfp583,Sppl2b,Faap24,Recql5,A230051N06Rik,Ints6l,Rnaseh2b,Strada,Scnn1a,Snhg17,Prelid3a,Grhl1,Icam4,Slc44a5,Tyw1,Parp8,Asb11,Vipr1,Zfp668 Ppp3ca,Nrgn,Tmsb4x,Actb,Snap25,Atp1b1,Olfm1,Ywhah,Calm2,Ncdn,Rtn1,Ptk2b,Calm1,Ppp3r1,Actg1,Chn1,Fth1,Ndrg4,Fkbp1a,Cfl1
2 101,255,101 0.116 523314 Cst3,Glul,Slc1a2,Mt1,Apoe,Sparcl1,Aldoc,Clu,Atp1a2,Slc1a3,Camk2a,Ndrg2,Gfap,Mt2,Gpr37l1,Atp1b2,Fam107a,Prdx6,Bc1,Plpp3 Pdk4,Bgn,Unc93b1,Rfx4,Ccr5,Phkg1,Slc38a3,Pax6,Chil1,Gli3,Thbs4,Ppp1r18,Sh3pxd2b,Lyz2,Cpt1a,Aif1,Pdlim4,Cd33,Lcat,Arhgef19 Cst3,Glul,Slc1a2,Sparcl1,Camk2a,Cpe,Mt1,Apoe,Mbp,Fth1,Aldoc,Clu,Ttr,Camk2n1,Ckb,Ddn,Rps29,Atp1a2,Scd2,Mt3
3 101,254,255 0.09284 418815 Fam163b,Prox1,Adcy1,Stxbp6,C1ql2,Btbd3,Synpr,Sema5a,Dsp,Eef1a1,Ncdn,Jun,Lrrtm4,Rfx3,Olfm1,Dgkh,Marcksl1,Ncald,Pitpnm2,Nrgn Il20rb,Tdo2,Col22a1,C1ql2,St3gal1,Plk5,Dsp,Prox1os,Fam163b,Prox1,Rph3al,H2bc6,Lrrtm4,Mcm6,Khdrbs2,Vwa3b,Prdm5,Npnt,Dact1,Stxbp6 Nrgn,Eef1a1,Ncdn,Olfm1,Ppp3ca,Actb,Tmsb4x,Calm1,Camk2a,Sparcl1,Atp1b1,Adcy1,Fam163b,Tspan7,Ndrg4,Rplp1,Arf3,Camk2n1,Rpl17,Ptk2b
4 101,101,255 0.03885 175240 Plp1,Mbp,Gatm,Mobp,Cnp,Cldn11,Fth1,Car2,Ermn,Cryab,Qdpr,Tubb4a,Trf,Plekhb1,Qki,Tspan2,Mal,Bcas1,Tmem88b,Septin4 Insc,Creb5,Tmem125,Trim36,Hapln2,Pde1c,Gjc2,Prr5l,Gjc3,Sec14l5,Gatm,Nkx6-2,Tmem88b,Adamts4,Plekhh1,Ermn,Plp1,Trim59,Plekhg3,Enpp6 Mbp,Plp1,Fth1,Mobp,Ptgds,Actb,Gatm,Tubb4a,Cnp,Car2,Cldn11,Tpt1,Scd2,Glul,Qdpr,App,Qki,Ptma,Malat1,Lars2
5 255,101,254 0.02347 105862 Ttr,Ptgds,Enpp2,Tac2,Gng8,Sostdc1,Ecrg4,Zic1,Calb2,Nnat,Dcn,Tmem212,Adcyap1,Gpr151,2900040C04Rik,Necab2,Apod,Nhlh2,Pou4f1,Ace Tmem212,Kcne2,Cldn2,Dcn,Septin10,Col8a2,Pou4f1,Nhlh2,Gng8,Wif1,Chrna3,Adcyap1,Folr1,Col1a2,Clec3b,Gng14,Sostdc1,Foxc1,Ecrg4,Tac2 Ttr,Ptgds,Enpp2,Apoe,Cpe,Pcp4,Actb,Eef1a1,Nnat,Psap,Fth1,Tpt1,Cst3,Dbi,Atp1b1,Cox8a,Sparcl1,Clu,Rpl23,Cox7c
6 255,178,101 0.00163 7366 Nkx2-2,Abtb2,Myo1d,C030029H02Rik,Pogk,Dusp16,Smco3,Gjb1,Ldlrad3,Rhobtb3,Lrrc8c,Cdr2,Piga,Tjap1,Carns1,Gpt,Prim1,Sh3gl3,Plk3,Cerox1 Nkx2-2,Abtb2,C030029H02Rik,Myo1d,Dusp16,Smco3,Gjb1,Pogk,Prim1,Ldlrad3,Plk3,Vgll4,Carns1,Gpt,Letm2,Cdr2,Piga,Fign,Lmln,Tjap1 Mbp,Fth1,Pogk,Malat1,Qdpr,Plp1,Myo1d,Tubb4a,Abtb2,Rps27a,Rhobtb3,Glul,Gatm,Nkx2-2,Sh3gl3,Alkbh5,C030029H02Rik,Smco3,Tpt1,Gjb1
7 178,255,101 0.00073 3290 Sst,Crhbp,Npy,Cort,Reln,Uhrf1bp1,Rab3b,2310010J17Rik,Elfn1,Lypd6b,Lgals1,Rpp25,Gad2,Cdh13,Clic5,Cenpf,Dlx1,Bcam,Foxred2,Sec14l5 Sst,Uhrf1bp1,Crhbp,Cort,Reln,Lypd6b,2310010J17Rik,Elfn1,Rab3b,Npy,Clic5,Rpp25,Cenpf,Bcam,Lgals1,Sec14l5,Dlx1,Foxred2,Gpc3,Cdh13 Sst,Npy,Crhbp,Atp1b1,Reln,Cort,Zwint,Gad2,Rab3b,Syt1,2310010J17Rik,Snap25,Sparcl1,Atp1a3,Oxr1,Scg2,Mdh1,Atp6v0c,Atp6v1e1,Vgf
9 101,178,255 0.0004 1817 Hba-a2,Hbb-bs,Hba-a1,Hbb-bt,Tent5c,Polr2l,Aven,Tinagl1,Map2k3,Rgs6,Bst2,Zfp318,Stk40,Plekha8,Ube2l6,Rad50,Slc12a4,Kank2,Srgap1,Klf2 Hba-a1,Hbb-bs,Hbb-bt,Hba-a2,Tent5c,Polr2l,Aven,Tinagl1,Map2k3,Rgs6,Bst2,Slc12a4,Stk40,Ube2l6,Plekha8,Kank2,Rad50,Zfp318,Prtg,Srgap1 Hba-a2,Hbb-bs,Hba-a1,Hbb-bt,Polr2l,Atp1b1,Camk2a,Plekhb1,Tent5c,Rpl31,Praf2,Tpd52,Ptgds,Ddn,Fth1,Calm3,Rpl38,Mobp,Psme3,Mkrn1
8 0,223,95 5e-05 214 Sass6,Recql5,Jmjd4,A230072C01Rik,Zfp583,Eef1akmt2,Rnaseh2b,Orc6,Akap10,Mak16,Slc25a35,Dclre1b,C130074G19Rik,Fbxw4,Arhgap10,Il17ra,Zbtb2,Katnip,Hook2,Ints6l Sass6,Recql5,Jmjd4,A230072C01Rik,Zfp583,Rnaseh2b,Eef1akmt2,Akap10,Orc6,Slc25a35,Dclre1b,C130074G19Rik,Mak16,Arhgap10,Il17ra,Fbxw4,Katnip,Zbtb2,Hook2,Ints6l Sass6,Recql5,Jmjd4,Cst3,Mak16,A230072C01Rik,Orc6,Tmsb4x,Zfp583,Hexb,Eef1akmt2,Capzb,Gabarapl1,Mdh2,Odc1,Rpl6,Olfm1,Rps8,Golga1,Zfp106

Packed SGE and Spatial Factor Outputs from run_cartload2

The packed SGE data and spatial factor inferences generated by FICTURE are available in PMTile format on Zenodo: DOI:10.5281/zenodo.15759403.

These datasets can also be loaded directly using the following catalog YAML file:
https://zenodo.org/records/15802634/catalog.yaml