Philipp Spohr
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Philipp Spohr has a total of 38 co-authors. Here are the top co-authors by number of shared publications:
- Gunnar W. Klau
Chair Algorithmic Bioinformatics, Heinrich Heine University Düsseldorf, Düsseldorf, Germany
8 shared publications - Birgit Henrich
Institute of Medical Microbiology and Hospital Hygiene, Heinrich Heine University Düsseldorf, University Hospital Düsseldorf, Düsseldorf, Germany
5 shared publications - Rainer Haas
Department of Hematology, Immunology, and Clinical Immunology, Heinrich Heine University Düsseldorf, University Hospital Düsseldorf
5 shared publications - Alexander Dilthey
Heinrich-Heine-University Düsseldorf
4 shared publications - Klaus Pfeffer
Institute of Medical Microbiology and Hospital Hygiene, Heinrich Heine University Düsseldorf, 40225 Düsseldorf, Germany
4 shared publications - Sebastian Alexander Scharf
Institute of Medical Microbiology and Hospital Hygiene, Heinrich Heine University Düsseldorf, University Hospital Düsseldorf, Düsseldorf, Germany
4 shared publications
Latest Publications
- Open Access
Barcode Crosstalk in ONT Multiplex Sequencing: Quantification and Mitigation Strategies
Multiplexing samples in long-read sequencing with Oxford Nanopore Next Generation Sequencing Technology (ONT) by ligating specific native barcodes to individual DNA samples enables significant increases of high throughput sequencing combined with a significant reduction of sequencing costs. However, this advantage carries the risk of barcode misassignment / crosstalk. Employing ONT multiplex sequencing with samples, we observed misassigned barcodes so called barcode crosstalk, after ONT library preparation according to the standard protocol, particularly in samples with low input DNA concentrations. We assumed that these barcode misassignments are largely due to misligation of remaining native barcodes during subsequent the subsequent sequencing adapter ligation. To systematically investigate and quantify barcode crosstalk, genomic DNA (gDNA) from four bacterial type strains with different DNA input concentrations was prepared using three protocols for library preparation: the Nanopore standard protocol (protocol A: version valid until July 2, 2025) the new Nanopore protocol (protocol B: version from July 2, 2025), and an in house protocol with pooling of the barcoded samples only after the sequencing adapter ligation step (protocol C: in house). All samples were sequenced on a Nanopore PromethIon device. The results clearly showed that the use of protocol A resulted in a pronounced barcode crosstalk especially detectable in samples with low DNA input concentrations (up to 2.4% misassigned reads). The ONT adjustment in protocol B (altered washing buffer vs. protocol A) significantly alleviated the barcode crosstalk to below 0.01%, whereas protocol C eliminated barcode crosstalk virtually completely. These observations emphasize that sequencing results obtained with older ONT native barcoding protocol variants should be critically reviewed. The newer ONT barcoding protocol is preferable for sequencing, but it does not completely eliminate the barcode crosstalk effect. In conclusion, for low DNA input and high accuracy sequencing, protocol C is recommended.
Open AccessBarcode Crosstalk in ONT Multiplex Sequencing: Quantification and Mitigation Strategies
Multiplexing samples in long-read sequencing with Oxford Nanopore Next Generation Sequencing Technology (ONT) by ligating specific native barcodes to individual DNA samples enables significant increases of high throughput sequencing combined with a significant reduction of sequencing costs. However, this advantage carries the risk of barcode misassignment / crosstalk. Employing ONT multiplex sequencing with samples, we observed misassigned barcodes so called barcode crosstalk, after ONT library preparation according to the standard protocol, particularly in samples with low input DNA concentrations. We assumed that these barcode misassignments are largely due to misligation of remaining native barcodes during subsequent the subsequent sequencing adapter ligation. To systematically investigate and quantify barcode crosstalk, genomic DNA (gDNA) from four bacterial type strains with different DNA input concentrations was prepared using three protocols for library preparation: the Nanopore standard protocol (protocol A: version valid until July 2, 2025) the new Nanopore protocol (protocol B: version from July 2, 2025), and an in house protocol with pooling of the barcoded samples only after the sequencing adapter ligation step (protocol C: in house). All samples were sequenced on a Nanopore PromethIon device. The results clearly showed that the use of protocol A resulted in a pronounced barcode crosstalk especially detectable in samples with low DNA input concentrations (up to 2.4% misassigned reads). The ONT adjustment in protocol B (altered washing buffer vs. protocol A) significantly alleviated the barcode crosstalk to below 0.01%, whereas protocol C eliminated barcode crosstalk virtually completely. These observations emphasize that sequencing results obtained with older ONT native barcoding protocol variants should be critically reviewed. The newer ONT barcoding protocol is preferable for sequencing, but it does not completely eliminate the barcode crosstalk effect. In conclusion, for low DNA input and high accuracy sequencing, protocol C is recommended.
- Open Access
- 4 August 2025
- Published by IOP Publishing in Machine Learning: Health
Dynamic Prediction of Mortality Risk Following Allogeneic Hematopoietic Stem Cell Transplantation
- 4 August 2025
- Published by IOP Publishing in Machine Learning: Health
Open AccessDynamic Prediction of Mortality Risk Following Allogeneic Hematopoietic Stem Cell Transplantation
- Open Access
Dynamic Mortality Risk Prediction in Myelodysplastic Syndromes Using Longitudinal Clinical Data
Purpose: Patients with myelodysplastic syndromes (MDS) exhibit diverse disease trajectories necessitating different clinical approaches ranging from watch-and-wait strategies to hematopoietic stem cell transplantation. Existing risk scores like the IPSS-R or EASIX provide static risk stratification at diagnosis but do not capture evolving disease dynamics. We address this problem by introducing a dynamic, data-driven approach to predict short-term mortality risks repeatedly, across the patient’s disease course. Materials and methods: We develop a machine learning model based on gradient-boosted decision trees to estimate one-year mortality risks from both longitudinal parameters from blood values and diagnosis-based features. We train the model on a dataset of patients from the MDS Registry Düsseldorf (n=1024) and validate on patients from University Hospitals Heidelberg (n=286) and Mannheim (n=31). Results: Validations on independent cohorts achieved AUROC scores of around 0.8 and better predictive performance for one-year survival compared to a diagnosis-only baseline model. The model accurately predicted mortality risks as early as within the first 90 days of diagnosis. Feature importance analysis revealed clinically plausible feature-label relations, supporting interpretability. Conclusion: This dynamic risk model enables continuous, individualized assessment of one-year mortality risk in MDS patients, offering a supplement to static scores used at diagnosis. Our results highlight the utility and importance of including longitudinal parameters in risk assessment analysis.
Open AccessDynamic Mortality Risk Prediction in Myelodysplastic Syndromes Using Longitudinal Clinical Data
Purpose: Patients with myelodysplastic syndromes (MDS) exhibit diverse disease trajectories necessitating different clinical approaches ranging from watch-and-wait strategies to hematopoietic stem cell transplantation. Existing risk scores like the IPSS-R or EASIX provide static risk stratification at diagnosis but do not capture evolving disease dynamics. We address this problem by introducing a dynamic, data-driven approach to predict short-term mortality risks repeatedly, across the patient’s disease course. Materials and methods: We develop a machine learning model based on gradient-boosted decision trees to estimate one-year mortality risks from both longitudinal parameters from blood values and diagnosis-based features. We train the model on a dataset of patients from the MDS Registry Düsseldorf (n=1024) and validate on patients from University Hospitals Heidelberg (n=286) and Mannheim (n=31). Results: Validations on independent cohorts achieved AUROC scores of around 0.8 and better predictive performance for one-year survival compared to a diagnosis-only baseline model. The model accurately predicted mortality risks as early as within the first 90 days of diagnosis. Feature importance analysis revealed clinically plausible feature-label relations, supporting interpretability. Conclusion: This dynamic risk model enables continuous, individualized assessment of one-year mortality risk in MDS patients, offering a supplement to static scores used at diagnosis. Our results highlight the utility and importance of including longitudinal parameters in risk assessment analysis.
- Open Access
Dynamic Prediction of Mortality Risk Following Allogeneic Hematopoietic Stem Cell Transplantation
Allogeneic hematopoietic stem cell transplantation (alloHSCT) is a potentially curative treatment for high-risk hematological malignancies, but early mortality within 100 days remains a significant challenge, affecting approximately 18% of patients. Identifying high-risk patients early is critical for timely interventions. While static risk scores like the Hematopoietic Cell Transplantation-Specific Comorbidity Index (HCT-CI) and Endothelial Activation and Stress Index (EASIX) are valuable, they do not adapt to patients’ clinical trajectories. Leveraging the increasing availability of digital medical datasets, we developed a dynamic risk stratification system based on a random forest approach to continuously predict mortality risk from day 0 to day 30 post-transplantation. Analyzing data from 847 patients undergoing alloHSCT between 2004 and 2019, our model achieved high classification performance, with AUROC scores continuously exceeding 0.8 from day 17, and effectively stratified patients into high, moderate, and low-risk groups. Feature importance analysis revealed a shift over time, with early predictions utilizing diverse variables and later stages relying more on inflammation-related bloodwork parameters. This dynamic approach surpasses static methods, demonstrating improved predictive performance and adaptability across treatment phases. To promote reproducibility and broader application, we provide open access to the software and accompanying bloodwork time-series dataset.
Open AccessDynamic Prediction of Mortality Risk Following Allogeneic Hematopoietic Stem Cell Transplantation
Allogeneic hematopoietic stem cell transplantation (alloHSCT) is a potentially curative treatment for high-risk hematological malignancies, but early mortality within 100 days remains a significant challenge, affecting approximately 18% of patients. Identifying high-risk patients early is critical for timely interventions. While static risk scores like the Hematopoietic Cell Transplantation-Specific Comorbidity Index (HCT-CI) and Endothelial Activation and Stress Index (EASIX) are valuable, they do not adapt to patients’ clinical trajectories. Leveraging the increasing availability of digital medical datasets, we developed a dynamic risk stratification system based on a random forest approach to continuously predict mortality risk from day 0 to day 30 post-transplantation. Analyzing data from 847 patients undergoing alloHSCT between 2004 and 2019, our model achieved high classification performance, with AUROC scores continuously exceeding 0.8 from day 17, and effectively stratified patients into high, moderate, and low-risk groups. Feature importance analysis revealed a shift over time, with early predictions utilizing diverse variables and later stages relying more on inflammation-related bloodwork parameters. This dynamic approach surpasses static methods, demonstrating improved predictive performance and adaptability across treatment phases. To promote reproducibility and broader application, we provide open access to the software and accompanying bloodwork time-series dataset.
- Open Access
- 24 May 2024
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 40 (6)
SWGTS—a platform for stream-based host DNA depletion
Motivation: Microbial sequencing data from clinical samples is often contaminated with human sequences, which have to be removed prior to sharing. Existing methods for human read removal, however, are applicable only after the target dataset has been retrieved in its entirety, putting the recipient at least temporarily in control of a potentially identifiable genetic dataset with potential implications under regulatory frameworks such as the GDPR. In some instances, the ability to carry out stream-based host depletion as part of the data transfer process may be preferable. Results: We present SWGTS, a client–server application for the transfer and stream-based host depletion of sequencing reads. SWGTS enforces a robust upper bound on the maximum amount of human genetic data from any one client held in memory at any point in time by storing all incoming sequencing data in a limited-size, client-specific intermediate processing buffer, and by throttling the rate of incoming data if it exceeds the speed of host depletion carried out on the SWGTS server in the background. SWGTS exposes a HTTP–REST interface, is implemented using docker-compose, Redis and traefik, and requires less than 8 Gb of RAM for deployment. We demonstrate high filtering accuracy of SWGTS; incoming data transfer rates of up to 1.65 megabases per second in a conservative configuration; and mitigation of re-identification risks by the ability to limit the number of SNPs present on a popular population-scale genotyping array covered by reads in the SWGTS buffer to a low user-defined number, such as 10 or 100. Availability and implementation: SWGTS is available on GitHub: https://github.com/AlBi-HHU/swgts (https://doi.org/10.5281/zenodo.10891052). The repository also contains a jupyter notebook that can be used to reproduce all the benchmarks used in this article. All datasets used for benchmarking are publicly available.
- 24 May 2024
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 40 (6)
Open AccessSWGTS—a platform for stream-based host DNA depletion
Motivation: Microbial sequencing data from clinical samples is often contaminated with human sequences, which have to be removed prior to sharing. Existing methods for human read removal, however, are applicable only after the target dataset has been retrieved in its entirety, putting the recipient at least temporarily in control of a potentially identifiable genetic dataset with potential implications under regulatory frameworks such as the GDPR. In some instances, the ability to carry out stream-based host depletion as part of the data transfer process may be preferable. Results: We present SWGTS, a client–server application for the transfer and stream-based host depletion of sequencing reads. SWGTS enforces a robust upper bound on the maximum amount of human genetic data from any one client held in memory at any point in time by storing all incoming sequencing data in a limited-size, client-specific intermediate processing buffer, and by throttling the rate of incoming data if it exceeds the speed of host depletion carried out on the SWGTS server in the background. SWGTS exposes a HTTP–REST interface, is implemented using docker-compose, Redis and traefik, and requires less than 8 Gb of RAM for deployment. We demonstrate high filtering accuracy of SWGTS; incoming data transfer rates of up to 1.65 megabases per second in a conservative configuration; and mitigation of re-identification risks by the ability to limit the number of SNPs present on a popular population-scale genotyping array covered by reads in the SWGTS buffer to a low user-defined number, such as 10 or 100. Availability and implementation: SWGTS is available on GitHub: https://github.com/AlBi-HHU/swgts (https://doi.org/10.5281/zenodo.10891052). The repository also contains a jupyter notebook that can be used to reproduce all the benchmarks used in this article. All datasets used for benchmarking are publicly available.
- Open Access
- 19 February 2024
- Published by Springer Nature in Scientific Reports
- Vol. 14 (1) , 1-20
Insights into gut microbiomes in stem cell transplantation by comprehensive shotgun long-read sequencing
The gut microbiome is a diverse ecosystem, dominated by bacteria; however, fungi, phages/viruses, archaea, and protozoa are also important members of the gut microbiota. Exploration of taxonomic compositions beyond bacteria as well as an understanding of the interaction between the bacteriome with the other members is limited using 16S rDNA sequencing. Here, we developed a pipeline enabling the simultaneous interrogation of the gut microbiome (bacteriome, mycobiome, archaeome, eukaryome, DNA virome) and of antibiotic resistance genes based on optimized long-read shotgun metagenomics protocols and custom bioinformatics. Using our pipeline we investigated the longitudinal composition of the gut microbiome in an exploratory clinical study in patients undergoing allogeneic hematopoietic stem cell transplantation (alloHSCT; n = 31). Pre-transplantation microbiomes exhibited a 3-cluster structure, characterized by Bacteroides spp. /Phocaeicola spp., mixed composition and Enterococcus abundances. We revealed substantial inter-individual and temporal variabilities of microbial domain compositions, human DNA, and antibiotic resistance genes during the course of alloHSCT. Interestingly, viruses and fungi accounted for substantial proportions of microbiome content in individual samples. In the course of HSCT, bacterial strains were stable or newly acquired. Our results demonstrate the disruptive potential of alloHSCTon the gut microbiome and pave the way for future comprehensive microbiome studies based on long-read metagenomics.
- 19 February 2024
- Published by Springer Nature in Scientific Reports
- Vol. 14 (1) , 1-20
Open AccessInsights into gut microbiomes in stem cell transplantation by comprehensive shotgun long-read sequencing
The gut microbiome is a diverse ecosystem, dominated by bacteria; however, fungi, phages/viruses, archaea, and protozoa are also important members of the gut microbiota. Exploration of taxonomic compositions beyond bacteria as well as an understanding of the interaction between the bacteriome with the other members is limited using 16S rDNA sequencing. Here, we developed a pipeline enabling the simultaneous interrogation of the gut microbiome (bacteriome, mycobiome, archaeome, eukaryome, DNA virome) and of antibiotic resistance genes based on optimized long-read shotgun metagenomics protocols and custom bioinformatics. Using our pipeline we investigated the longitudinal composition of the gut microbiome in an exploratory clinical study in patients undergoing allogeneic hematopoietic stem cell transplantation (alloHSCT; n = 31). Pre-transplantation microbiomes exhibited a 3-cluster structure, characterized by Bacteroides spp. /Phocaeicola spp., mixed composition and Enterococcus abundances. We revealed substantial inter-individual and temporal variabilities of microbial domain compositions, human DNA, and antibiotic resistance genes during the course of alloHSCT. Interestingly, viruses and fungi accounted for substantial proportions of microbiome content in individual samples. In the course of HSCT, bacterial strains were stable or newly acquired. Our results demonstrate the disruptive potential of alloHSCTon the gut microbiome and pave the way for future comprehensive microbiome studies based on long-read metagenomics.
- Open Access
CATHI: An interactive platform for comparative genomics and homolog identification
A bstract: Bioinformatics has established itself as a central pillar of modern biology. Specifically, comparative genomics enables scientists to study a vast number of genomes efficiently. These comparative analyses shed light on the evolution and potential function of genomes and genes, but are also increasingly used as a key tool for metabolic engineering and synthetic biology by identifying appropriate targets for modification. While numerous sophisticated tools for comparative genomics and homolog identification exist, those tools predominantly target highly skilled bioinformatics users. Consequently, many biologists either defer such analyses to their more versed bioinformatic collaborators or resort to suboptimal tools. Here, we present an intuitive solution available on all major operating systems, easily accessed through common web browsers. CATHI – Comparative Analysis Tool for Homolog Identification – integrates a suite of best-practice bioinformatic tools, encompassing BLAST for homology searches, MAFFT for multiple sequence alignment, FastTree2 for phylogeny reconstruction, and clinker for synteny analysis. Specifically tailored to biologists, CATHI orchestrates predefined settings and automated pipelines, obviating the need for programming expertise. This platform empowers researchers to confidently engage in detailed comparative genomics studies by streamlining the analytical process. The interactive framework provides users with a plethora of options. This includes real-time execution and progress monitoring, facilitates dynamic result tracking, and a set of search functions across NCBI databases like CDD or ProtFam. Users can interactively engage in data exploration, filtering, and visualization through CATHI’s intuitive interface. Furthermore, the seamless export of project data in standard formats (FASTA, Newick, CSV, and HTML) facilitates the integration with further third-party tools such as TreeViewer and Jalview. To benchmark CATHI, we revisited the comparative analysis of cyanobacterial circadian clock proteins conducted by Schmelling et al. in 2017, revealing consistent global patterns among identified homologs, while also highlighting individual variations attributed to the expansion of available databases.
Open AccessCATHI: An interactive platform for comparative genomics and homolog identification
A bstract: Bioinformatics has established itself as a central pillar of modern biology. Specifically, comparative genomics enables scientists to study a vast number of genomes efficiently. These comparative analyses shed light on the evolution and potential function of genomes and genes, but are also increasingly used as a key tool for metabolic engineering and synthetic biology by identifying appropriate targets for modification. While numerous sophisticated tools for comparative genomics and homolog identification exist, those tools predominantly target highly skilled bioinformatics users. Consequently, many biologists either defer such analyses to their more versed bioinformatic collaborators or resort to suboptimal tools. Here, we present an intuitive solution available on all major operating systems, easily accessed through common web browsers. CATHI – Comparative Analysis Tool for Homolog Identification – integrates a suite of best-practice bioinformatic tools, encompassing BLAST for homology searches, MAFFT for multiple sequence alignment, FastTree2 for phylogeny reconstruction, and clinker for synteny analysis. Specifically tailored to biologists, CATHI orchestrates predefined settings and automated pipelines, obviating the need for programming expertise. This platform empowers researchers to confidently engage in detailed comparative genomics studies by streamlining the analytical process. The interactive framework provides users with a plethora of options. This includes real-time execution and progress monitoring, facilitates dynamic result tracking, and a set of search functions across NCBI databases like CDD or ProtFam. Users can interactively engage in data exploration, filtering, and visualization through CATHI’s intuitive interface. Furthermore, the seamless export of project data in standard formats (FASTA, Newick, CSV, and HTML) facilitates the integration with further third-party tools such as TreeViewer and Jalview. To benchmark CATHI, we revisited the comparative analysis of cyanobacterial circadian clock proteins conducted by Schmelling et al. in 2017, revealing consistent global patterns among identified homologs, while also highlighting individual variations attributed to the expansion of available databases.
- Open Access
MetaGut: Insights into gut microbiomes in stem cell transplantation by comprehensive shotgun long-read sequencing
The gut microbiome is a diverse ecosystem, dominated by bacteria; however, fungi, phages/viruses, archaea, and protozoa are also important members of the gut microbiota. Up to recently, exploration of taxonomic compositions beyond bacteria as well as an understanding of the interaction between the bacteriome with the other members was limited due to 16S rDNA sequencing. Here, we developed MetaGut, a method enabling the simultaneous interrogation of the gut microbiome (bacteriome, mycobiome, archaeome, eukaryome, DNA virome) and of antibiotic resistance genes based on optimized long-read shotgun metagenomics protocols and custom bioinformatics. Using MetaGut we investigated the longitudinal composition of the gut microbiome in an exploratory clinical study in patients undergoing allogeneic hematopoietic stem cell transplantation (alloHSCT; n = 31). Pre-transplantation microbiomes exhibited a 3-cluster structure, associated with Bacteroides / Phocaeicola , mixed composition and Enterococcus abundances. MetaGut revealed substantial inter-individual and temporal variabilities of microbial domain compositions, human DNA, and antibiotic resistance genes during the course of alloHSCT. Interestingly, viruses and fungi accounted for substantial proportions of microbiome content in individual samples (up to >50% and >20%, respectively). After leukopenia, strains were stable or newly acquired. Our results demonstrate the disruptive effect of alloHSCT on the gut microbiome and pave the way for future studies based on long-read metagenomics.
Open AccessMetaGut: Insights into gut microbiomes in stem cell transplantation by comprehensive shotgun long-read sequencing
The gut microbiome is a diverse ecosystem, dominated by bacteria; however, fungi, phages/viruses, archaea, and protozoa are also important members of the gut microbiota. Up to recently, exploration of taxonomic compositions beyond bacteria as well as an understanding of the interaction between the bacteriome with the other members was limited due to 16S rDNA sequencing. Here, we developed MetaGut, a method enabling the simultaneous interrogation of the gut microbiome (bacteriome, mycobiome, archaeome, eukaryome, DNA virome) and of antibiotic resistance genes based on optimized long-read shotgun metagenomics protocols and custom bioinformatics. Using MetaGut we investigated the longitudinal composition of the gut microbiome in an exploratory clinical study in patients undergoing allogeneic hematopoietic stem cell transplantation (alloHSCT; n = 31). Pre-transplantation microbiomes exhibited a 3-cluster structure, associated with Bacteroides / Phocaeicola , mixed composition and Enterococcus abundances. MetaGut revealed substantial inter-individual and temporal variabilities of microbial domain compositions, human DNA, and antibiotic resistance genes during the course of alloHSCT. Interestingly, viruses and fungi accounted for substantial proportions of microbiome content in individual samples (up to >50% and >20%, respectively). After leukopenia, strains were stable or newly acquired. Our results demonstrate the disruptive effect of alloHSCT on the gut microbiome and pave the way for future studies based on long-read metagenomics.
- Open Access
The Homo-Edit Distance Problem
We consider the homo-edit distance problem, which is the minimum number of homo-deletions or homo-insertions to convert one string into another. A homo-insertion is the insertion of a string of equal characters into another string, while a homo-deletion is the inverse operation. We show how to compute the homo-edit distance of two strings in polynomial time: We first demonstrate that the problem is equivalent to computing a common subsequence of the two input strings with a minimum number of homo-deletions and then present a dynamic programming solution for the reformulated problem. 2012 ACM Subject Classification: Applied computing → Bioinformatics; Applied computing → Molecular sequence analysis; Theory of computation → Dynamic programming
Open AccessThe Homo-Edit Distance Problem
We consider the homo-edit distance problem, which is the minimum number of homo-deletions or homo-insertions to convert one string into another. A homo-insertion is the insertion of a string of equal characters into another string, while a homo-deletion is the inverse operation. We show how to compute the homo-edit distance of two strings in polynomial time: We first demonstrate that the problem is equivalent to computing a common subsequence of the two input strings with a minimum number of homo-deletions and then present a dynamic programming solution for the reformulated problem. 2012 ACM Subject Classification: Applied computing → Bioinformatics; Applied computing → Molecular sequence analysis; Theory of computation → Dynamic programming