example bedfile has unexpected extra column with "JUN::JUNB"

[jwokaty]

When I download the bedfile associated with the following data (it's the example bedfile; I use it for most of the examples in bedbaser), although"bed_compliance": "bed4+0", and "non_compliant_columns": 0,, I see the following in the first few lines:

chr1	1000896	1000906	-1.000000e+11	JUN::JUNB
chr1	1066881	1066891	-1.000000e+11	JUN::JUNB
chr1	1208146	1208156	-1.000000e+11	JUN::JUNB
chr1	1218006	1218016	-1.000000e+11	JUN::JUNB
chr1	1318385	1318395	-1.000000e+11	JUN::JUNB

I think this must be an error processing because I expect 4 standard bedfile columns and no nonstandard columns from the metadata. Below is the metadata that I see for comparison.

{
  "name": "ENCFF445QIM",
  "genome_alias": "hg38",
  "genome_digest": "Ba88PY52_qeifhJrgUXyin6UITdXNsg3",
  "bed_compliance": "bed4+0",
  "data_format": "ucsc_bed",
  "compliant_columns": 4,
  "non_compliant_columns": 0,
  "header": null,
  "id": "2511272115fa7786e8bc8808b6c3d7de",
  "description": null,
  "submission_date": "2026-04-11T01:36:50.629086Z",
  "last_update_date": "2026-04-14T18:48:03.926378Z",
  "is_universe": false,
  "license_id": "DUO:0000042",
  "annotation": {
    "species_name": "Homo sapiens",
    "species_id": "9606",
    "genotype": "",
    "phenotype": "",
    "description": "",
    "cell_type": null,
    "cell_line": "Caki2",
    "tissue": "kidney, epithelium",
    "library_source": null,
    "assay": "DNase-seq",
    "antibody": null,
    "target": null,
    "treatment": null,
    "global_sample_id": [
      "encode:ENCFF445QIM"
    ],
    "global_experiment_id": [
      "encode:ENCSR825NXL"
    ],
    "original_file_name": "ENCFF445QIM.bed.gz"
  },
  "processed": false,
  "stats": null,
  "plots": null,
  "files": null,
  "universe_metadata": null,
  "raw_metadata": null,
  "bedsets": null
}

Let me know if I should file another place. Thanks for your help.

[jwokaty]

I see the bed file has changed since my original comment but the last line isn't complete for bed id = b35ab6dd84b827f916e2b6c96c2076b4:

chr1	37673849	37674342	AG93479	50.65625
chr1	37674643	37674

[khoroshevskyi]

Sorry, I didn't notice this issue earlier.

hmm, it seems a very big bug in processing of the bed files. And I found another file that was incorrectly processed:
2892a16d9b29bb7c24eccc97fa23b681

metadata:

Species Name	Homo sapiens
Species Id	9606
Cell Type	anterior cingulate cortex
Tissue	brain
Assay	DNase-seq
Global Sample Id	[encode:ENCFF003NPD](https://www.encodeproject.org/files/ENCFF003NPD)
Global Experiment Id	[encode:ENCSR586WGL](https://www.encodeproject.org/experiments/ENCSR586WGL)
Original File Name	ENCFF003NPD.bed.gz

[khoroshevskyi]

Related to:
databio/gtars#263

[khoroshevskyi]

we will try to fix this issue as fast as we can, but it can take couple of weeks 🤕

[jwokaty]

@khoroshevskyi No worries. Thanks for looking into it and let me know if I can help!

[khoroshevskyi]

I believe that I fixed all corrupted files in the database. Because files have changed, also their ids changed:

b35ab6dd84b827f916e2b6c96c2076b4: https://bedbase.org/bed/950c80d2a609847064779db835fe344e
2892a16d9b29bb7c24eccc97fa23b681: https://bedbase.org/bed/5df700f171f678e1f46c4a83a4443463

This files are not in the vector database yet, but I will add soon.

Let me know if you find any other incorrect files, or anything else that needs to be fixed:D

[khoroshevskyi]

I found another incorrect file: https://bedbase.org/bed/54ab15870fcf6623b9ff0a62ffd37b61

[jwokaty]

@khoroshevskyi Thanks for working on this.

I think there are still some issues. For example a0169b30e33ceaaf1614a83f56094727 has a 5th column but the metadata says it should have 4 compliant columns and no noncompliant columns:

> ex$id
[1] "a0169b30e33ceaaf1614a83f56094727"
> ex$bed_compliance
[1] "bed4+0"
> ex$non_compliant_columns
[1] 0

head a0169b30e33ceaaf1614a83f56094727.bed
chr1	1079580	1079595	-2.850659e+04	ESR2
chr1	1080387	1080402	-2.127936e+04	ESR2
chr1	1615177	1615192	-2.995155e+04	ESR2
chr1	2229256	2229271	-3.860696e+04	ESR2
chr1	2255936	2255951	-2.701756e+04	ESR2
chr1	2301595	2301610	-2.991022e+04	ESR2
chr1	2411948	2411963	-1.907262e+04	ESR2
chr1	2643407	2643422	-1.415884e+04	ESR2
chr1	3666638	3666653	-1.156741e+04	ESR2
chr1	3677424	3677439	-2.675737e+04	ESR2

Same with 0aa14c284a35cbdf4cdf0daea638e9b0.bed:

head  0aa14c284a35cbdf4cdf0daea638e9b0.bed
chr1	3186773	3186788	-4.674671e+04	SOX21
chr1	5975457	5975472	-4.495585e+04	SOX21
chr1	8343572	8343587	-5.046889e+04	SOX21
chr1	9798861	9798876	-2.542640e+04	SOX21
chr1	11012150	11012165	-4.178717e+04	SOX21
chr1	12435898	12435913	-4.011102e+04	SOX21
chr1	14304567	14304582	-6.732980e+04	SOX21
chr1	16639498	16639513	-2.612863e+04	SOX21
chr1	16727518	16727533	-4.182000e+04	SOX21
chr1	19505573	19505588	-2.960856e+04	SOX21

I see that when i look at them at bedbase.org, I see the sign that they haven't been processed by bedboss. What does this mean?

[khoroshevskyi]

@jwokaty sorry for the late reply. Yes I saw this issue, but I didn't determine where the actual problem is yet. All this tasks are on my todo list 🫣

[khoroshevskyi]

@jwokaty
Answering your questions:

1. If it sais that bed file haven't been processed by bedboss, it means that some of the statistics weren't calculated yet, but file, metadata, and classification metadata should be processed correctly.
2. This file that you provided a0169b30e33ceaaf1614a83f56094727, is corrupted file and on the line 1655:

chr18	48456605	48456616	-nan	ESR2

it has -nan value, which brakes pandas, and leads to this weird result. I will try to fix it, but there are very few files that have this problem, and overall very unique problem