The 2015 MiTranscriptome study (Iyer et al., Nature Genetics 47, 199-208) reconstructed a consensus human transcriptome from 7,256 RNA-seq libraries and cataloged 58,648 long noncoding RNA (lncRNA) genes plus a separate class of transcripts of unknown coding potential (TUCPs), nominating 7,942 lncRNA or TUCP genes (11,478 transcripts) as lineage- or cancer-associated, while leaving a connected set of open tasks that still define the central gap of the lncRNA field. First, only 31% of RefSeq transcripts were recovered with their exact splicing patterns, so complete full-length isoform structures were (and remain) unavailable for most lncRNA loci, monoexonic and intragenic loci being hardest. Second, coding-potential assignments were provisional: TUCPs were flagged from sequence features alone, and some annotated lncRNAs appear to be partial fragments of larger protein-coding genes, anticipating that tumor-derived proteomics would settle these cases. Third, essentially all disease nominations rested on computational evidence, with no causal mechanisms, in vivo models, or eQTL links between disease-associated variants and lncRNA expression. Work since has advanced, but not closed, each strand: long-read sequencing consortia (LRGASP) and GENCODE's capture long-read program substantially advanced structural completeness without closing it; ribosome profiling and tumor immunopeptidomics detected thousands of translated noncanonical open reading frames while reference annotation still deliberately withholds protein-coding status for them; genome-scale CRISPR interference screens, documented perturbation-method confounds, and the first human Mendelian disorder caused by a lncRNA-locus deletion (CHASERR) mark real but sparse functional progress. The problem - complete structures, settled coding potential, and causal functional validation for the cataloged lncRNA landscape - remains open.

Background

Long noncoding RNAs (lncRNAs) are RNA transcripts longer than about 200 nucleotides that are transcribed, like messenger RNAs, by RNA polymerase II but do not encode proteins; they regulate gene expression through mechanisms including chromatin modulation, enhancer-associated activity, and post-transcriptional control, and they have been repeatedly implicated in cancer. In 2015, Iyer and colleagues assembled the MiTranscriptome catalog by ab initio transcriptome reconstruction - computational reassembly of transcript structures directly from RNA sequencing (RNA-seq) reads, without relying on existing gene annotations - applied to 7,256 libraries from tumors, normal tissues, and cell lines. The result was 91,013 expressed genes, of which 58,648 were classified as lncRNAs - far outnumbering the entries in lncRNA databases at the time (fewer than 16,000 genes; the comparison catalogs GENCODE and RefSeq are curated community reference gene sets) - and 7,942 lncRNA or transcripts-of-unknown-coding-potential (TUCP) genes (11,478 transcripts) were nominated as lineage-specific or cancer-associated by a rank-based differential expression procedure. The catalog's own measurements exposed three connected weaknesses. (1) Structural incompleteness: although the assembly recovered 94% of RefSeq transcript nucleotides and 93% of annotated splice junctions, only 31% of RefSeq transcripts were reconstructed with their exact splicing pattern - the precise combination of exons defining an isoform - a long-standing limitation of short-read assembly that is aggravated for monoexonic (single-exon) and intragenic (transcribed within another gene) loci, which earlier catalogs had often avoided annotating altogether. (2) Uncertain coding potential: alongside the lncRNA class the authors defined transcripts of unknown coding potential (TUCPs), long RNAs flagged as possibly protein-coding by the Coding Potential Assessment Tool (CPAT, an alignment-free logistic-regression classifier using open reading frame size and sequence composition) or by a Pfam protein-domain match. Some cataloged lncRNAs, exemplified by the LINC00514 locus that turned out to contain a 29-exon open reading frame of 418 amino acids, appeared to be partial fragments of larger protein-coding genes, and only 268 TUCP genes found peptide support in a benign-tissue proteomics dataset, prompting the authors to anticipate that tumor-derived proteomics would strengthen the predictions. (3) Functional untestedness: the disease and lineage nominations were computational; only 100 transcripts were validated by RT-PCR, and no causal mechanisms, in vivo disease models, or expression quantitative trait locus (eQTL, a genetic variant statistically associated with the expression level of a transcript) links connecting genome-wide association study (GWAS) disease variants to lncRNA expression existed for the new entries. These weaknesses jointly define what later literature calls the catalog-to-function gap: without complete transcript structures, functional assays can target fragmentary or wrong models; without settled coding potential, 'noncoding' mechanisms may in fact be protein effects; and without experimental causality testing, disease nominations remain correlations. Ribosome profiling (Ribo-seq, sequencing of ribosome-protected RNA fragments that maps translated open reading frames) and CRISPR interference (CRISPRi, a nuclease-dead Cas9 system that represses transcription of a targeted locus without cutting DNA) are the key technologies later work brought to bear on the gap.

Problem Statement

Close the catalog-to-function gap for the human long noncoding RNA transcriptome that the 2015 MiTranscriptome landscape study exposed, taking as its object the catalog that study produced (58,648 lncRNA genes and the accompanying TUCP class, of which 7,942 lncRNA or TUCP genes - 11,478 transcripts - were nominated as lineage- or cancer-associated). The gap comprises three linked problems: the first two range over the full catalog (the 58,648 lncRNA genes and the accompanying TUCP class), while the third ranges over the 7,942 nominated lineage- or cancer-associated lncRNA and TUCP genes. (1) Structural completeness: determine the complete full-length transcript structures - all expressed isoforms with their exact splicing patterns - at every cataloged lncRNA locus, including the monoexonic and intragenic loci that short-read assembly resolves least reliably. (2) Coding-potential resolution: assign each cataloged lncRNA or TUCP-class transcript an evidence-based coding-status determination - protein-coding, noncoding, or translated with unresolved function where translation is detected but the product's biological role is not established - that distinguishes genuine noncoding RNAs from partially annotated or unannotated protein-coding genes and replaces the provisional sequence-feature predictions of 2015. (3) Functional and disease causality: experimentally establish which of the nominated lineage- or cancer-associated lncRNA and TUCP genes have causal roles in normal physiology and disease, including their molecular mechanisms, and determine how disease-associated GWAS variants in regions proximal to these loci relate to lncRNA expression and function. The three problems are interdependent - a functional assay of an incompletely structured or actually protein-coding transcript is uninterpretable, and an artifact catalog cannot be functionally annotated - and jointly convert a computational catalog into a reference-grade, functionally interpreted annotation of the human noncoding transcriptome. Clinical biomarker development for the nominated lncRNAs is the translational consequence of problem (3).

References

  1. Iyer MK, Niknafs YS, Malik R, Singhal U, Sahu A, Hosono Y, Barrette TR, Prensner JR, Evans JR, Zhao S, Poliakov A, Cao X, Dhanasekaran SM, Wu YM, Robinson DR, Beer DG, Feng FY, Iyer HK, Chinnaiyan AM. The landscape of long noncoding RNAs in the human transcriptome. Nature Genetics 47, 199-208 (2015). DOI: 10.1038/ng.3192. https://doi.org/10.1038/ng.3192 (open-access full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC4417758/)
  2. Uszczynska-Ratajczak B, Lagarde J, Frankish A, Guigó R, Johnson R. Towards a complete map of the human long non-coding RNA transcriptome. Nature Reviews Genetics 19, 535-548 (2018). DOI: 10.1038/s41576-018-0017-y. https://doi.org/10.1038/s41576-018-0017-y
  3. Lorenzi L, Chiu HS, Avila Cobos F, Gross S, Volders PJ, Cannoodt R, Nuytens J, Vanderheyden K, Anckaert J, Lefever S, Tay AP, de Bony EJ, Trypsteen W, Gysens F, Vromman M, Goovaerts T, Hansen TB, Kuersten S, Nijs N, Taghon T, Vermaelen K, Bracke KR, Saeys Y, De Meyer T, Deshpande NP, Anande G, Chen TW, Wilkins MR, Unnikrishnan A, De Preter K, Kjems J, Koster J, Schroth GP, Vandesompele J, Sumazin P, Mestdagh P. The RNA Atlas expands the catalog of human non-coding RNAs. Nature Biotechnology 39, 1453-1465 (2021). DOI: 10.1038/s41587-021-00936-1. https://doi.org/10.1038/s41587-021-00936-1
  4. Pardo-Palacios FJ, Wang D, Reese F, Diekhans M, Carbonell-Sala S, Williams B, Loveland JE, De María M, Adams MS, Balderrama-Gutierrez G, Behera AK, Gonzalez Martinez JM, Hunt T, Lagarde J, Liang CE, Li H, Meade MJ, Moraga Amador DA, Prjibelski AD, Birol I, Bostan H, Brooks AM, Çelik MH, Chen Y, Du MRM, Felton C, Göke J, Hafezqorani S, Herwig R, Kawaji H, Lee J, Li JL, Lienhard M, Mikheenko A, Mulligan D, Nip KM, Pertea M, Ritchie ME, Sim AD, Tang AD, Wan YK, Wang C, Wong BY, Yang C, Barnes I, Berry AE, Capella-Gutierrez S, Cousineau A, Dhillon N, Fernandez-Gonzalez JM, Ferrández-Peral L, Garcia-Reyero N, Götz S, Hernández-Ferrer C, Kondratova L, Liu T, Martinez-Martin A, Menor C, Mestre-Tomás J, Mudge JM, Panayotova NG, Paniagua A, Repchevsky D, Ren X, Rouchka E, Saint-John B, Sapena E, Sheynkman L, Smith ML, Suner MM, Takahashi H, Youngworth IA, Carninci P, Denslow ND, Guigó R, Hunter ME, Maehr R, Shen Y, Tilgner HU, Wold BJ, Vollmers C, Frankish A, Au KF, Sheynkman GM, Mortazavi A, Conesa A, Brooks AN (Long-read RNA-seq Genome Annotation Assessment Project Consortium). Systematic assessment of long-read RNA-seq methods for transcript identification and quantification. Nature Methods 21, 1349-1363 (2024). DOI: 10.1038/s41592-024-02298-3. https://doi.org/10.1038/s41592-024-02298-3
  5. Mudge JM, Carbonell-Sala S, Diekhans M, Martinez JG, Hunt T, Jungreis I, Loveland JE, Arnan C, Barnes I, Bennett R, Berry A, Bignell A, Cerdán-Vélez D, Cochran K, Cortés LT, Davidson C, Donaldson S, Dursun C, Fatima R, Hardy M, Hebbar P, Hollis Z, James BT, Jiang Y, Johnson R, Kaur G, Kay M, Mangan RJ, Maquedano M, Gómez LM, Mathlouthi N, Merritt R, Ni P, Palumbo E, Perteghella T, Pozo F, Raj S, Sisu C, Steed E, Sumathipala D, Suner MM, Uszczynska-Ratajczak B, Wass E, Yang YT, Zhang D, Finn RD, Gerstein M, Guigó R, Hubbard TJP, Kellis M, Kundaje A, Paten B, Tress ML, Birney E, Martin FJ, Frankish A. GENCODE 2025: reference gene annotation for human and mouse. Nucleic Acids Research 53, D966-D975 (2025). DOI: 10.1093/nar/gkae1078. https://doi.org/10.1093/nar/gkae1078
  6. Han SW, Jewell S, Thomas-Tikhonenko A, Barash Y. Contrasting and combining transcriptome complexity captured by short and long RNA sequencing reads. Genome Research 34, 1624-1635 (2024). DOI: 10.1101/gr.278659.123. https://doi.org/10.1101/gr.278659.123
  7. Liu SJ, Horlbeck MA, Cho SW, Birk HS, Malatesta M, He D, Attenello FJ, Villalta JE, Cho MY, Chen Y, Mandegar MA, Olvera MP, Gilbert LA, Conklin BR, Chang HY, Weissman JS, Lim DA. CRISPRi-based genome-scale identification of functional long noncoding RNA loci in human cells. Science 355, aah7111 (2017). DOI: 10.1126/science.aah7111. https://doi.org/10.1126/science.aah7111
  8. Horlbeck MA, Liu SJ, Chang HY, Lim DA, Weissman JS. Fitness effects of CRISPR/Cas9-targeting of long noncoding RNA genes. Nature Biotechnology 38, 573-576 (2020). DOI: 10.1038/s41587-020-0428-0. https://doi.org/10.1038/s41587-020-0428-0
  9. Prensner JR, Enache OM, Luria V, Krug K, Clauser KR, Dempster JM, Karger A, Wang L, Stumbraite K, Wang VM, Botta G, Lyons NJ, Goodale A, Kalani Z, Fritchman B, Brown A, Alan D, Green T, Yang X, Jaffe JD, Roth JA, Piccioni F, Kirschner MW, Ji Z, Root DE, Golub TR. Noncanonical open reading frames encode functional proteins essential for cancer cell survival. Nature Biotechnology 39, 697-704 (2021). DOI: 10.1038/s41587-020-00806-2. https://doi.org/10.1038/s41587-020-00806-2
  10. Ouspenskaia T, Law T, Clauser KR, Klaeger S, Sarkizova S, Aguet F, Li B, Christian E, Knisbacher BA, Le PM, Hartigan CR, Keshishian H, Apffel A, Oliveira G, Zhang W, Chen S, Chow YT, Ji Z, Jungreis I, Shukla SA, Justesen S, Bachireddy P, Kellis M, Getz G, Hacohen N, Keskin DB, Carr SA, Wu CJ, Regev A. Unannotated proteins expand the MHC-I-restricted immunopeptidome in cancer. Nature Biotechnology 40, 209-217 (2022). DOI: 10.1038/s41587-021-01021-3. https://doi.org/10.1038/s41587-021-01021-3
  11. Statello L, Guo CJ, Chen LL, Huarte M. Gene regulation by long non-coding RNAs and its biological functions. Nature Reviews Molecular Cell Biology 22, 96-118 (2021). DOI: 10.1038/s41580-020-00315-9. https://doi.org/10.1038/s41580-020-00315-9
  12. Mattick JS, Amaral PP, Carninci P, Carpenter S, Chang HY, Chen LL, Chen R, Dean C, Dinger ME, Fitzgerald KA, Gingeras TR, Guttman M, Hirose T, Huarte M, Johnson R, Kanduri C, Kapranov P, Lawrence JB, Lee JT, Mendell JT, Mercer TR, Moore KJ, Nakagawa S, Rinn JL, Spector DL, Ulitsky I, Wan Y, Wilusz JE, Wu M. Long non-coding RNAs: definitions, functions, challenges and recommendations. Nature Reviews Molecular Cell Biology 24, 430-447 (2023). DOI: 10.1038/s41580-022-00566-8. https://doi.org/10.1038/s41580-022-00566-8
  13. Ganesh VS, Riquin K, Chatron N, Yoon E, Lamar KM, Aziz MC, Monin P, O'Leary MC, Goodrich JK, Garimella KV, England E, Weisburd B, Aguet F, Bacino CA, Murdock DR, Dai H, Rosenfeld JA, Emrick LT, Ketkar S, Sarusi Y, Sanlaville D, Kayani S, Broadbent B, Pengam A, Isidor B, Bezieau S, Cogné B, MacArthur DG, Ulitsky I, Carvill GL, O'Donnell-Luria A. Neurodevelopmental Disorder Caused by Deletion of CHASERR, a lncRNA Gene. New England Journal of Medicine 391, 1511-1518 (2024). DOI: 10.1056/NEJMoa2400718. https://doi.org/10.1056/NEJMoa2400718
  14. Bennett M, Muitjens S, Ulitsky I, Baker AH. In search of LncDACH1 - a perceived intronic long non-coding RNA lacking fundamental genomic annotation evidence. Nature Communications 17 (2026). DOI: 10.1038/s41467-026-76991-6. https://doi.org/10.1038/s41467-026-76991-6
  15. Bennett M, Ulitsky I, Baker AH. Missing 'lncs': how insufficient scrutiny can lead to misrepresentation of long non-coding RNAs (lncRNAs) and their function in cardiovascular disease. European Heart Journal 46, 3434-3436 (2025). DOI: 10.1093/eurheartj/ehaf115. https://doi.org/10.1093/eurheartj/ehaf115
  16. Li Z, Liu L, Feng C, Qin Y, Xiao J, Zhang Z, Ma L. LncBook 2.0: integrating human long non-coding RNAs with multi-omics annotations. Nucleic Acids Research 51, D186-D191 (2023). DOI: 10.1093/nar/gkac999. https://doi.org/10.1093/nar/gkac999