Zsolt Balazs
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Zsolt Balazs has a total of 166 co-authors. Here are the top co-authors by number of shared publications:
- Michael Krauthammer
Department of Quantitative Biomedicine, University of Zurich, Zurich, Switzerland
141 shared publications - Aizhan Tastanova
Department of Dermatology, University Hospital Zurich, Zurich, Switzerland
136 shared publications - Patrick Turko
Department of Dermatology, University of Zurich, University Hospital Zurich, Zurich, Switzerland
135 shared publications - Mitchell Paul Levesque
Department of Dermatology, University of Zurich, University Hospital Zurich, Zurich, Switzerland
134 shared publications - Reinhard Dummer
Department of Dermatology, University Hospital Zürich, University of Zurich, 8091 Zurich, Switzerland
134 shared publications - Alaa Othman
ETH Zurich
133 shared publications
Latest Publications
- Open Access
- 11 May 2026
- Published by Elsevier in SSRN Electronic Journal
Epigenetic Modifications during Pregnancy in Patients with Rheumatoid Arthritis and Systemic Lupus Erythematosus: A Multicentre Study
Background: Increasing number of epigenetic variations have been characterized in rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE).Aim: The aim of this pilot study was to prospectively evaluate pregnancy and postpartum-related epigenetic, hormonal and immunologic modifications in RA/SLE patients, versus healthy women (HS).Methods: A total of 46 pregnant (15 HS, 13 RA, 18 SLE) and 50 non-pregnant women (15 HS, 16 RA, 19 SLE) were enrolled in 4 Centres (Genoa, Brescia, Milan, Ghent). Clinical data collection, hormonal (human chorionic gonadotropin (hCG), estradiol, progesterone, prolactin), interleukin (IL)-10 dosages and epigenetic analyses (DNA methylation, chromatin accessibility/ ATAC-seq) were performed at each trimester and postpartum. DNA methylation data were used to estimate the deconvoluted distribution of different immune cells.Findings: Estradiol and progesterone serum levels showed a statistically significant increase during pregnancy in all study groups, whereas prolactin levels were significantly higher in HS and SLE pregnant vs non-pregnant women (p<0.001 and p<0.01, respectively). Pregnancy induced most differences in chromatin accessibility in the post-partum period, also in HS also in the first trimester. In HS, changes in DNA methylation occurred mainly in the first and second trimesters, with gradual normalization over the course of pregnancy. In RA, pregnancy induced fewer differentially methylated sites compared to non-pregnant, with most changes occurring after delivery. SLE pregnancies showed the lowest amount of DNA methylation changes among all groups, with most changes occurring after delivery. Differently to RA, in SLE patients changes were shifted toward hypomethylation compared to non-pregnant patients. Interpretation: This is the first study to prospectively report pregnancy-related epigenetic changes in RA and SLE patients compared to HS, suggesting a coordinated, disease-specific pattern, potentially involved in hormonal-immune modulation of pregnancy outcomes.
- 11 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessEpigenetic Modifications during Pregnancy in Patients with Rheumatoid Arthritis and Systemic Lupus Erythematosus: A Multicentre Study
Background: Increasing number of epigenetic variations have been characterized in rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE).Aim: The aim of this pilot study was to prospectively evaluate pregnancy and postpartum-related epigenetic, hormonal and immunologic modifications in RA/SLE patients, versus healthy women (HS).Methods: A total of 46 pregnant (15 HS, 13 RA, 18 SLE) and 50 non-pregnant women (15 HS, 16 RA, 19 SLE) were enrolled in 4 Centres (Genoa, Brescia, Milan, Ghent). Clinical data collection, hormonal (human chorionic gonadotropin (hCG), estradiol, progesterone, prolactin), interleukin (IL)-10 dosages and epigenetic analyses (DNA methylation, chromatin accessibility/ ATAC-seq) were performed at each trimester and postpartum. DNA methylation data were used to estimate the deconvoluted distribution of different immune cells.Findings: Estradiol and progesterone serum levels showed a statistically significant increase during pregnancy in all study groups, whereas prolactin levels were significantly higher in HS and SLE pregnant vs non-pregnant women (p<0.001 and p<0.01, respectively). Pregnancy induced most differences in chromatin accessibility in the post-partum period, also in HS also in the first trimester. In HS, changes in DNA methylation occurred mainly in the first and second trimesters, with gradual normalization over the course of pregnancy. In RA, pregnancy induced fewer differentially methylated sites compared to non-pregnant, with most changes occurring after delivery. SLE pregnancies showed the lowest amount of DNA methylation changes among all groups, with most changes occurring after delivery. Differently to RA, in SLE patients changes were shifted toward hypomethylation compared to non-pregnant patients. Interpretation: This is the first study to prospectively report pregnancy-related epigenetic changes in RA and SLE patients compared to HS, suggesting a coordinated, disease-specific pattern, potentially involved in hormonal-immune modulation of pregnancy outcomes.
- Open Access
LoRTIA Plus: a chemistry-agnostic, feature-first software package for long-read transcriptome annotation
Long-read RNA sequencing (lrRNA-seq) enables direct reconstruction of full-length transcripts, yet existing annotation tools show variable performance across genomes and library chemistries, particularly for novel isoforms. We present LoRTIA Plus, a chemistry-agnostic suite for transcriptome annotation and reconstruction from lrRNA-seq data. LoRTIA Plus first detects and filters transcription start sites (TSSs), transcription end sites (TESs), and introns using adapter-aware and quality-based criteria, and evaluates read support before assembling high-confidence transcript models. We benchmarked LoRTIA Plus against bambu, FLAIR, IsoQuant, and NAGATA on KSHV transcriptomes with dense overlap, using a validated literature-supported boundary set, and on transcriptomes from three human cell lines from the Long Read RNA-seq Genome Annotation Assessment Project (LRGASP) sequenced with five long-read chemistries. On KSHV, LoRTIA Plus achieved the highest F1 scores for TSSs, TESs, and transcripts in both direct-cDNA and direct-RNA datasets by improving recall without sacrificing precision. Across human datasets, LoRTIA Plus consistently ranked among the top boundary annotators across all chemistries and was the best-performing tool in PCR-based libraries, while remaining highly competitive on native RNA. Junction- and isoform-level analyses show that LoRTIA Plus yields a rich, reproducible repertoire of novel isoforms and transcript boundaries from viral to human transcriptomes.
Open AccessLoRTIA Plus: a chemistry-agnostic, feature-first software package for long-read transcriptome annotation
Long-read RNA sequencing (lrRNA-seq) enables direct reconstruction of full-length transcripts, yet existing annotation tools show variable performance across genomes and library chemistries, particularly for novel isoforms. We present LoRTIA Plus, a chemistry-agnostic suite for transcriptome annotation and reconstruction from lrRNA-seq data. LoRTIA Plus first detects and filters transcription start sites (TSSs), transcription end sites (TESs), and introns using adapter-aware and quality-based criteria, and evaluates read support before assembling high-confidence transcript models. We benchmarked LoRTIA Plus against bambu, FLAIR, IsoQuant, and NAGATA on KSHV transcriptomes with dense overlap, using a validated literature-supported boundary set, and on transcriptomes from three human cell lines from the Long Read RNA-seq Genome Annotation Assessment Project (LRGASP) sequenced with five long-read chemistries. On KSHV, LoRTIA Plus achieved the highest F1 scores for TSSs, TESs, and transcripts in both direct-cDNA and direct-RNA datasets by improving recall without sacrificing precision. Across human datasets, LoRTIA Plus consistently ranked among the top boundary annotators across all chemistries and was the best-performing tool in PCR-based libraries, while remaining highly competitive on native RNA. Junction- and isoform-level analyses show that LoRTIA Plus yields a rich, reproducible repertoire of novel isoforms and transcript boundaries from viral to human transcriptomes.
- Open Access
Reconstructing real-world metastatic lines of therapy enables progression risk stratification in breast cancer
Estimating progression-free survival (PFS) for each metastatic line of therapy (mLoT) is valuable yet challenging in precision oncology, as both are rarely available as paired labels. We present (i) an evidence enrichment pipeline that reconstructs mLoTs and PFS from longitudinal EHR and (ii) a model for progression risk prognostication. In the MSK-CHORD breast cancer cohort, the pipeline yielded 8,791 mLoTs from 2,881 patients, and the trained model achieved Antolini’s C = 0.681 ± 0.006 and integrated Brier score (IBS) = 0.124 ± 0.004 in nested cross-validation. We validated our pipeline by reconstructing regimen and PFS labels in the AACR GENIE BPC cohorts from DFCI and VICC. Of the reconstructed regimens, 86.2% and 82.4% matched a curated regimen, with PFS event agreement 93.2% and 94.6% among matches. On the curated BPC regimens, the MSK-CHORD model achieved C = 0.649 ± 0.002 and 0.634 ± 0.006 , with IBS = 0.166 ± 0.001 and 0.159 ± 0.001 , despite several features of original model missing. The model remains robust to modality ablation and stratifies prognosis interpretably despite substantial cohort imbalance and heterogeneity. Together, this supports scalable characterization of metastatic-line PFS across heterogeneous regimens and hypothesis generation, pending prospective clinical validation.
Open AccessReconstructing real-world metastatic lines of therapy enables progression risk stratification in breast cancer
Estimating progression-free survival (PFS) for each metastatic line of therapy (mLoT) is valuable yet challenging in precision oncology, as both are rarely available as paired labels. We present (i) an evidence enrichment pipeline that reconstructs mLoTs and PFS from longitudinal EHR and (ii) a model for progression risk prognostication. In the MSK-CHORD breast cancer cohort, the pipeline yielded 8,791 mLoTs from 2,881 patients, and the trained model achieved Antolini’s C = 0.681 ± 0.006 and integrated Brier score (IBS) = 0.124 ± 0.004 in nested cross-validation. We validated our pipeline by reconstructing regimen and PFS labels in the AACR GENIE BPC cohorts from DFCI and VICC. Of the reconstructed regimens, 86.2% and 82.4% matched a curated regimen, with PFS event agreement 93.2% and 94.6% among matches. On the curated BPC regimens, the MSK-CHORD model achieved C = 0.649 ± 0.002 and 0.634 ± 0.006 , with IBS = 0.166 ± 0.001 and 0.159 ± 0.001 , despite several features of original model missing. The model remains robust to modality ablation and stratifies prognosis interpretably despite substantial cohort imbalance and heterogeneity. Together, this supports scalable characterization of metastatic-line PFS across heterogeneous regimens and hypothesis generation, pending prospective clinical validation.
- Open Access
- 8 February 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Uncovering Cas9 PAM diversity through metagenomic mining and machine learning
Recognition of protospacer adjacent motifs (PAMs) is crucial for target site recognition by CRISPR–Cas systems. In genome editing applications, the requirement for specific PAM sequences at the target locus imposes substantial constraints, driving efforts to search for novel Cas9 orthologs with extended or alternative PAM compatibilities. Here, we present CRISPR-PAMdb, a comprehensive and publicly accessible database compiling Cas9 protein sequences from 3.8 million bacterial and archaeal genomes and PAM profiles from 7.4 million phage and plasmid sequences. Through spacer–protospacer alignment, we infer consensus PAM preferences for 8003 unique Cas9 clusters. To extend PAM discovery beyond traditional alignment-based approaches, we develop CICERO, a machine learning model predicting PAM preferences directly from Cas9 protein sequences. Built on the ESM2 protein language model and trained on the CRISPR–PAMdb database, CICERO achieves an average cosine similarity of 0.69 on test data and 0.75 on experimentally validated Cas9 orthologs. For Cas9 clusters where alignment-based predictions are infeasible, CICERO generates PAM profiles for an additional 50,308 Cas9 proteins, including 17,453 high-confidence predictions with CICERO confidence scores above 0.8. Together, CRISPR–PAMdb and CICERO enable large-scale exploration of PAM diversity across Cas9 proteins, accelerating design of next-generation CRISPR-Cas9 tools for precise genome engineering.
- 8 February 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessUncovering Cas9 PAM diversity through metagenomic mining and machine learning
Recognition of protospacer adjacent motifs (PAMs) is crucial for target site recognition by CRISPR–Cas systems. In genome editing applications, the requirement for specific PAM sequences at the target locus imposes substantial constraints, driving efforts to search for novel Cas9 orthologs with extended or alternative PAM compatibilities. Here, we present CRISPR-PAMdb, a comprehensive and publicly accessible database compiling Cas9 protein sequences from 3.8 million bacterial and archaeal genomes and PAM profiles from 7.4 million phage and plasmid sequences. Through spacer–protospacer alignment, we infer consensus PAM preferences for 8003 unique Cas9 clusters. To extend PAM discovery beyond traditional alignment-based approaches, we develop CICERO, a machine learning model predicting PAM preferences directly from Cas9 protein sequences. Built on the ESM2 protein language model and trained on the CRISPR–PAMdb database, CICERO achieves an average cosine similarity of 0.69 on test data and 0.75 on experimentally validated Cas9 orthologs. For Cas9 clusters where alignment-based predictions are infeasible, CICERO generates PAM profiles for an additional 50,308 Cas9 proteins, including 17,453 high-confidence predictions with CICERO confidence scores above 0.8. Together, CRISPR–PAMdb and CICERO enable large-scale exploration of PAM diversity across Cas9 proteins, accelerating design of next-generation CRISPR-Cas9 tools for precise genome engineering.
- Open Access
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Table ST2 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
Table ST2: Summary of the clinical data corresponding to the primary cell cultures used in this publication
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Open AccessTable ST2 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
Table ST2: Summary of the clinical data corresponding to the primary cell cultures used in this publication
- Open Access
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Supplementary Data from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
Tumor Profiler Consortium author list
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Data from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
Tumor Profiler Consortium author list
- Open Access
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Table ST6 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
Table ST6: Differential expressed lipids and their FDRs in sensitive vs resistant cells (negative values are upregulated in sensitive cells).
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Open AccessTable ST6 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
Table ST6: Differential expressed lipids and their FDRs in sensitive vs resistant cells (negative values are upregulated in sensitive cells).
- Open Access
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Suppl. Figure S8 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Open AccessSuppl. Figure S8 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
- Open Access
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Suppl Figure S4 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
S4: Neocuproine is a copper-dependent ROS inducer
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Open AccessSuppl Figure S4 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
S4: Neocuproine is a copper-dependent ROS inducer
- Open Access
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Suppl. Figure S1 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
S1: Evaluation of HST screening.
- 24 November 2025
- Published by American Association for Cancer Research (AACR)
Open AccessSuppl. Figure S1 from ROS Induction Targets Persister Cancer Cells with Low Metabolic Activity in NRAS-Mutated Melanoma
S1: Evaluation of HST screening.