# paper.pdf ## Metadata - PDFFormatVersion=1.6 - Language=en-US - IsLinearized=true - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - CreationDate=D:20260702111912+05'30' - Creator=Adobe InDesign 17.4 (Windows) - ModDate=D:20260702125201+05'30' - Producer=Adobe PDF Library 16.0.7 - Trapped=False - xmpmm:instanceid=uuid:c3aa70ce-354e-4b05-8d00-84637a5fef2a - xmpmm:originaldocumentid=xmp.did:A33916180F2068118083ACD1B377E157 - xmpmm:documentid=xmp.id:19d79eb0-ddf1-8540-926f-d5dc69be64fa - xmpmm:renditionclass=proof:pdf - xmpmm:derivedfrom=xmp.iid:a3422f4f-6cec-9240-8fac-88e6e0f8d7dcxmp.did:a3422f4f-6cec-9240-8fac-88e6e0f8d7dcxmp.did:A33916180F2068118083ACD1B377E157default - xmpmm:history=convertedfrom application/x-indesign to application/pdfAdobe InDesign 17.4 (Windows)/2026-07-02T11:19:12+05:30 - xmp:createdate=2026-07-02T11:19:12+05:30 - xmp:modifydate=2026-07-02T12:52:01+05:30 - xmp:metadatadate=2026-07-02T12:52:01+05:30 - xmp:creatortool=Adobe InDesign 17.4 (Windows) - xmprights:marked=False - photoshop:modifydate=2019-01-31T16:16:25 - photoshop:createdate=2019-01-24T11:43:46 - dc:format=application/pdf - pdf:producer=Adobe PDF Library 16.0.7 - pdf:trapped=False - extensisfontsense:slug=3023672065GraphikOpenType - PS3023672065--0Graphik-Bold1.0011904271029GraphikOpenType - PS1904271029--0Graphik-Regular1.0012428528031News Gothic MT OTOpenType - PS2428528031Monotype Typography0NewsGothicMTOT-Regular1.000543760319News Gothic MT OTOpenType - PS543760319Monotype Typography0NewsGothicMTOT-Bold1.0002785939051KnockoutOpenType - PS2785939051--0Knockout-68FullFeatherwt1.200423856041Graphik X CondOpenType - PS423856041--0GraphikXCond-Super1.0001526035435Graphik X CondOpenType - PS1526035435--0GraphikXCond-Bold1.0003093728306Graphik CondOpenType - PS3093728306--0GraphikCond-Black1.0011341118539Graphik CondOpenType - PS1341118539--0GraphikCond-Regular1.0011385893531GraphikOpenType - PS1385893531--0Graphik-Semibold1.0012986479693Graphik CompactOpenType - PS2986479693--0GraphikCompact-Regular1.0024223138361Graphik CompactOpenType - PS4223138361--0GraphikCompact-Semibold1.0023673047546Harding SevenOpenType - PS3673047546--0HardingSeven-Extrabold0.000103712566Harding SevenOpenType - PS103712566--0HardingSeven-Roman0.0002232452979Harding SevenOpenType - PS2232452979--0HardingSeven-Medium0.000631916388Harding SevenOpenType - PS631916388--0HardingSeven-Bold0.000157629484Publico BannerOpenType - PS157629484--0PublicoBanner-Black1.005 ## Contents ### Page 1 nature metabolismhttps://doi.org/10.1038/s42255-026-01543-7ArticleA data-driven framework reconstructs the molecular continuum of human MASLD progressionIn the format provided by the authors and uneditedSupplementary information ### Page 2 R = 0.99, p < 2.2e−16 0 5 10 15 20 0 5 10 15 20 Normalised CTRL (SANYAL) Corrected CTRL (SANYAL) CTRL SANYALR = 0.98, p < 2.2e−16 0 5 10 15 20 0 5 10 15 20 Normalised MASL (SANYAL) Corrected MASL (SANYAL) MASL SANYALR = 0.98, p < 2.2e−16 0 5 10 15 20 0 5 10 15 20 Normalised MASHF0 (SANYAL) Corrected MASHF0 (SANYAL) MASHF0 SANYALR = 0.98, p < 2.2e−16 0 5 10 15 20 0 5 10 15 20 Normalised MASHF1 (SANYAL) Corrected MASHF1 (SANYAL) MASHF1 SANYALR = 0.98, p < 2.2e−16 0 5 10 15 20 0 5 10 15 20 Normalised MASHF2 (SANYAL) Corrected MASHF2 (SANYAL) MASHF2 SANYALR = 0.98, p < 2.2e−16 0 5 10 15 20 0 5 10 15 20 Normalised MASHF34 (SANYAL) Corrected MASHF34 (SANYAL) MASHF34 SANYALBefore/After Batch Correction (normalised using 1st column)SANYAL_CTRL SANYAL_MASL SANYAL_F0 SANYAL_F1 SANYAL_F2 SANYAL_F3 SANYAL_F4 UCAM_MASL UCAM_F1 UCAM_F2 UCAM_F3 UCAM_F4 SANYAL_CTRL* SANYAL_MASL* SANYAL_F0* SANYAL_F1* SANYAL_F2* SANYAL_F3* SANYAL_F4* UCAM_MASL* UCAM_F1* UCAM_F2* UCAM_F3* UCAM_F4* MMP9 MMP2 COL1A1 COL3A1 COL4A2 TIMP1 COL1A2 COL4A1 ACTA2 TIMP2 TGFB1 COL6A1Diet_Group D0.6 0.8 1 1.2 1.4A B C DSANYAL UCAM UCAM + SANYAL−50 0 50 100−100 −80 −60 −40 −20 0 20 40(Merged UCAM & Sanyal dataset)PC1 − 55.5 % PC2 − 3.6 % CTRL MASL MASHF012 MASHF34−100 −50 0−40 −20 0 20 40 60(Merged UCAM & Sanyal dataset)PC1 − 9 % PC2 − 4.9 % CTRL MASL MASHF012 MASHF34Sample_5 (Outlier) Sample_5 (Outlier)Before/After Batch Correction (normalised using 1st column)MMP9 MMP2 COL1A1 COL3A1 COL4A2 TIMP1 COL1A2 COL4A1 ACTA2Diet_Group Diet_GroupSANYAL_before_correction UCAM_before_correction SANYAL_after_correction UCAM_after_correction 0.8 1 1.2 1.4R = 0.97, p < 2.2e−16 0 5 10 15 20 5 10 15 20 Normalised MASL (UCAM) Corrected MASL (UCAM) MASL UCAMR = 0.97, p < 2.2e−16 0 5 10 15 20 5 10 15 20 Normalised MASHF1 (UCAM) Corrected MASHF1 (UCAM) MASHF1 UCAMR = 0.97, p < 2.2e−16 0 5 10 15 20 5 10 15 20 Normalised MASHF2 (UCAM) Corrected MASHF2 (UCAM) MASHF2 UCAMR = 0.97, p < 2.2e−16 0 5 10 15 20 5 10 15 20 Normalised MASHF3 (UCAM) Corrected MASHF3 (UCAM) MASHF3 UCAMR = 0.97, p < 2.2e−16 0 5 10 15 20 5 10 15 20 Normalised MASHF4 (UCAM) Corrected MASHF4 (UCAM) MASHF4 UCAME F G0.4 0.6 0.8 1.0 Low Medium High Expression level Mean Pearson correlation ### Page 3 Supplementary Figure 1. Quality control and validation for dataset integration. A. PCA plot of the merged UCAM/VCU quantile normalised counts. There is an apparent batch effect separating the patients into two clusters stemming from the dataset from which they are derived. B. COMBAT batch effect correction solved the problem of the batch effect with patients altogether in a cluster and a clear disease progression pattern from the right towards the later disease stages on the left. The outlier (sample 5) was removed from all downstream analyses. C-D. Normalised counts before and after batch effect correction, separately for each disease stage. The data suggest that COMBAT has not significantly altered gene expression, as indicated by the high correlation (R > 0.97) across all MASLD stages in the VCU (C) and UCAM (D) datasets. Inset in panel D shows the mean correlation in expression before and after correction stratified by low, medium, and high expression levels. The categories were defined by ranking genes by normalised expression in each stage and dividing them into three groups: High (top third), Medium (middle third), and Low (bottom third). The dots inside each bar correspond to the correlations of expression separately for each disease stage (dot colours in agreement with disease stage colours from panel C). E. Correlation of the principal components (PCs) with the variables of interest. The first 10 PCs have been correlated using Pearson correlation with 6 continuous variables of interest: age, NAS score, Fibrosis, Steatosis, Ballooning, and Inflammation. F. Heatmap of the quantile normalised counts before and after the merging/batch effect correction for UCAM and VCU datasets. The rows show genes previously linked to MASLD120–123. The data are shown in a “fold change vs the first disease stage” of each dataset, respectively (“red” = increase; “blue” = decrease; “white” = unchanged; “green datasets”: UCAM and VCU before correction; “blue datasets”: UCAM and VCU after correction; the asterisks “*” in the column names indicate that the samples have been batch-effect corrected). Note that the control group (the first column in each block) in the VCU dataset is labelled as CTRL, whereas in the UCAM it is labelled as MASL.