BUSCADOR RECOLECTA

This dataset was intended to describe schistosomiasis severity in a backcross cohort and to study the genetic linkage analysis with parasitological, pathological and immunological variables., [Description of methods used for collection/generation of data] F1BX mice were infected with 150 ± 5 S. mansoni cercariae each mouse and nine weeks post-infection were euthanized. We considered 20 variables: granulomas; affected liver surface (mm2/cm2); the number of adult male and female worms; eggs per gram of liver and small intestine; eggs in liver and small intestine per female; CD4, CD8, CD45, CD220 in blood or spleen; IgG, IgG1, IgG2a, IgM antibodies. Multivariate models (cluster and principal component analyses and K-means) identified four levels of infection intensity in the cohort. The genetic regions associated with severity were assessed by massive genotyping and genetic linkage analysis using 961 informative SNPs., [Methods for processing the data] Mean and standard error in each variable, Kolmogorov-Smirnov test. ANOVA and Tukey’ test or Student t-test. The Pearson correlation coefficient (r) and Student t-test for the statistical significance. Multivariant models considering sex influence in worm recovery, eggs in the liver and intestine, fecundity, granulomas and the affected liver surface. All the variables were standardized to 0 mean and standard deviation to 1. Cluster analysis, dendrogram, Principal components analysis and conglomerates by k-means were used to generate clusters. Median proportions were performed. SIMFIT statistical package for Windows version 7.3.7 were used Massive genotyping and geneticlinkage analysis using 961 informative SNPs: The genetic distance based on the recombination frequencies between markers in the F1BX cohort was compared with http://cgd.jax.org/mousemapconverter using maximum-likelihood mapping with HM algorithm. The Haldane function was used with a step size of 2.5 cM and a genotyping error of 0.001. We used the LOD-score to calculate the statistical significance of the linkage of the QTLs found. LOD-score higher than 1.4 suggested linkage. The Ensembl bioinformatics tool (https://www.ensembl.org/index.html) was used to identify syntenic regions between mouse and human., Here we present the dataset used in our study entitled "Identification of genomic regions implicated in susceptibility to Schistosoma mansoni infection in a murine genetic model (backcross)". Thus, we crossed the C57BL/6 mouse strain with the CBA one and then the F1B6CBA females (C57 x CBA) were backcrossed with CBA males generating the F1BX cohort of the study. The study consists of the identification of genetic markers of schistosomiasis. High infection levels and severe liver fibrosis in schistosomiasis appear only in a few highly susceptible infected people. Schistosomiasis could be a complex trait disease and it could be possible to identify genetic markers associated with severity. This study uses a genetically heterogeneous back-cross cohort with genetically unique mice. A backcross (F1BX) mouse cohort was generated after two stages; firstly, we crossed a mouse strain (CBA/2J) susceptible to schistosomiasis with a resistant one (C57BL/6J) to generate the F1B6CBA mice; secondly, the F1BX mice were generated by backcrossing. F1B6CBA female mice with CBA/2J males. F1BX mice were infected with 150 ± 5 S. mansoni cercariae each mouse and nine weeks post-infection were euthanized. We considered 20 variables: granulomas; affected liver surface (mm2/cm2); the number of adult male and female worms; eggs per gram of liver and small intestine; eggs in liver and small intestine per female; CD4, CD8, CD45, CD220 in blood or spleen; IgG, IgG1, IgG2a, IgM antibodies. Multivariate models (cluster and principal component analyses and K-means) identified four levels of infection intensity in the cohort. The genetic regions associated with severity were assessed by massive genotyping and genetic linkage analysis using 961 informative SNPs., The main research project is: Red de Investigación Colaborativa en Enfermedades Tropicales (RICET) Ref.: RD16/0027/0018., Peer reviewed

Supplementary Table 1. Characteristics of samples in Trp53-null mammary tumor cohort., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 2. Characteristics of samples in MMTV-Erbb2 mammary tumor cohort., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 3. Characteristics of patients in the TCGA-BRCA cohort., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 6. Univariate Cox proportional hazards regression result of CMB on overall survival in the TCGA-BRCA cohort., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 7. Defferentially expressed genes between Subtype 2 and Subtype 1 patients., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 10. Gene ontology (GO) functional enrichment analysis of the differentially expressed genes (DEGs) for molecular function., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 11. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis on differentially expressed genes (DEGs)., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Table 12. Cox proportional hazards regression analysis on the overall survival of patients in the TCGA-BRCA cohort., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed

Supplementary Figure 1. Representative examples of 256 CMB learned from Trp53-null mouse mammary tumors. Supplementary Figure 2. Consensus clustering on the Trp53-null mouse mammary tumors with different number of clusters (K) and the corresponding Kaplan–Meier curves for tumor growth. A-B. Consensus matrix with 3 and 4 clusters, respectively; C-D Kaplan–Meier curves for 3 and 4 subtypes, respectively. Supplementary Figure 3. Representative example of CMB_13 (A), CMB_249 (D), CMB_120 (G), and CMB_105 (J), and their significant and consistent difference in relative abundance between metastasis ground truth (B, E, H, and K) and low/high metastasis risk groups (i.e., LMRG and HMRG defined by CMS-1 and CMS-2, respectively) (C, F, I, and L). Supplementary Figure 4. BRCA patient subtypes in triple-negative (TNBC) and non-triplenegative (Non-TNBC) groups. A-B. KM curves for representative CMBs show consistent and significant impact on OS in Non-TNBC and TNBC groups, respectively; C. Subtype-specific patients in TCGA-BRCA cohort form distinct clusters in patient-level cellular morphometric context space in Non-TNBC and TNBC groups, respectively; D. Subtype-specific patients in TCGA-BRCA cohort show significant difference in survival in Non-TNBC and TNBC groups, respectively. Supplementary Figure 5. A. BRCA patient heatmap with mouse CMS model on the TCGABRCA cohort; B. BRCA patient heatmap with BC-CMS model on the TCGA-BRCA cohort. C. ROC curves for the prediction of 5-,10-, and 20-year overall survival of BRCA patients using all significant prognostic factors as listed in E; D. Comparison of predictive power between BC-CMS model and mouse CMS model using bootstrapping strategy with 80% sampling rate and 1000 iterations; E. Similar to patient subtype from BC-CMS model as shown in Figure 3F, patient subtype directly predicted from the mouse CMS model is also a significant and independent prognostic factor in the TCGA-BRCA cohort. Supplementary Figure 6. BC-CMS in triple-negative (TNBC) and non-triple-negative (NonTNBC) groups in the TCGA-BRCA cohort show significant difference in tumor microenvironments., Mouse models of cancer provide a powerful tool for investigating all aspects of cancer biology. In this study, we used our recently developed machine learning approach to identify the cellular morphometric biomarkers (CMB) from digital images of hematoxylin and eosin (H&E) micrographs of orthotopic Trp53-null mammary tumors (n = 154) and to discover the corresponding cellular morphometric subtypes (CMS). Of the two CMS identified, CMS-2 was significantly associated with shorter survival (p = 0.0084). We then evaluated the learned CMB and corresponding CMS model in MMTV-Erbb2 transgenic mouse mammary tumors (n = 53) in which CMS-2 was significantly correlated with the presence of metastasis (p = 0.004). We next evaluated the mouse CMB and CMS model on The Cancer Genome Atlas breast cancer (TCGA-BRCA) cohort (n = 1017). Kaplan–Meier analysis showed significantly shorter overall survival (OS) of CMS-2 patients compared to CMS-1 patients (p = 0.024) and added significant prognostic value in multi-variable analysis of clinical and molecular factors, namely, age, pathological stage, and PAM50 molecular subtype. Thus, application of CMS to digital images of routine workflow H&E preparations can provide unbiased biological stratification to inform patient care., Peer reviewed