Biology subjects

Kromp, F.

Publications and source records attributed to Kromp, F..

2 recordsLinked to original sources

bioRxiv2020

Single-cell landscape of bone marrow metastases in human neuroblastoma unraveled by deep multiplex imaging

Bone marrow commonly serves as a metastatic niche for disseminated tumor cells (DTCs) of solid cancers in patients with unfavorable clinical outcome. Single-cell assessment of bone marrow metastases is essential to decipher the entire spectrum of tumor heterogeneity in these cancers, however, has previously not been performed. Here we used multi-epitope-ligand cartography (MELC) to spatially profile 20 biomarkers and assess morphology in DTCs as well as hematopoietic and mesenchymal cells of eight bone marrow metastases from neuroblastoma patients. We developed DeepFLEX, a single-cell image analysis pipeline for MELC data that combines deep learning-based cell and nucleus segmentation and overcomes frequent challenges of multiplex imaging methods including autofluorescence and unspecific antibody binding. Using DeepFLEX, we built a single-cell atlas of bone marrow metastases comprising more than 35,000 single cells. Comparisons of cell type proportions between samples indicated that microenvironmental changes in the metastatic bone marrow are associated with tumor cell infiltration and therapy response. Hierarchical clustering of DTCs revealed multiple phenotypes with highly diverse expression of markers such as FAIM2, an inhibitory protein in the Fas apoptotic pathway, which we propose as a complementary marker to capture DTC heterogeneity in neuroblastoma. The presented single-cell atlas provides first insights into the heterogeneity of human bone marrow metastases and is an important step towards a deeper understanding of DTCs and their interactions with the bone marrow niche.

cancer biology↗

bioRxiv2018

Deep SNP: An End-to-end Deep Neural Network with Attention-based Localization for Break-point Detection in SNP Array Genomic data

Diagnosis and risk stratification of cancer and many other diseases require the detection of genomic breakpoints as a prerequisite of calling copy number alterations (CNA). This, however, is still challenging and requires time-consuming manual curation. As deep-learning methods outperformed classical state-of-the-art algorithms in various domains and have also been successfully applied to life science problems including medicine and biology, we here propose Deep SNP, a novel Deep Neural Network to learn from genomic data. Specifically, we used a manually curated dataset from 12 genomic single nucleotide polymorphism array (SNPa) profiles as truth-set and aimed at predicting the presence or absence of genomic breakpoints, an indicator of structural chromosomal variations, in windows of 40,000 probes. We compare our results with well-known neural network models as well as Rawcopy though this tool is designed to predict breakpoints and in addition genomic segments with high sensitivity. We show, that Deep SNP is capable of successfully predicting the presence or absence of a breakpoint in large genomic windows and outperforms state-of-the-art neural network models. Qualitative examples suggest that integration of a localization unit may enable breakpoint detection and prediction of genomic segments, even if the breakpoint coordinates were not provided for network training. These results warrant further evaluation of DeepSNP for breakpoint localization and subsequent calling of genomic segments.

genomics↗