Dr. Javad Alizargar • PD Dr. Carlo Maj • Philipps-University Marburg
Open review
Data live from DB
Workspace
Scientific foundations
Literature & audit
Call alertNext Humboldt call: 15 November 2026•Volume cap is reported as first 800 applications; verify on the official AvH portal before planning submission.
Track: Experienced Researcher•Host: Carlo Maj confirmed
Humboldt Research Fellowship · D201
Reproducibility and Perturbation Specificity of Genetic Regulatory Responses to DNA Damage in Human Immune Cells
A host-aligned research programme combining primary human blood work in Marburg with statistical genetics: response-eQTL discovery in CD8+ T cells, viability-aware interaction modelling, independent replication, and disease-GWAS colocalization.
Dr. Javad Alizargar, MD, PhD
Assistant Professor, National Taipei University of Nursing and Health Sciences, Taiwan
PhD 2019 · AvH eligibility confirmed · D201
PD Dr. Carlo Maj
Head of Bioinformatics & Research Group Leader, Center for Human Genetics, Philipps-University Marburg
Hosting agreement, workplace and English working language confirmed
Library records
—
DOI-deduplicated records; screening pending
Included papers
—
— screened / included
Search families
—
Logged search executions
Evidence records
—
Structured extractions; review status shown
Power at N = 70
—
β = 0.35, α = 1e-6; see calculator
Proposal draft
—
Words across live sections
Integrity correction applied 6 Nov 2026. The previous dashboard stated “84.2% power at N = 70.” That claim is not reproducible from the paired interaction model. The same inputs now yield less than 1% power at α = 1e-6, and 80% power would require an interaction effect of roughly 1.7 SD per allele. The proposal has been reframed as prespecified discovery plus independent replication. See the before quoting any power number.
1
Choose the scientific direction
Compare the four candidate topics on feasibility, novelty and Humboldt fit. Scores are internal prioritisation heuristics, not validated metrics. Carlo Maj has endorsed Topic A as the primary direction.
2
Separate facts from assumptions
Track what the host has confirmed, what is a proposed method, and what remains unresolved. The register now labels seeded notes as workspace notes and marks the interaction model as a proposal, not a confirmed fact.
3
Revise a defensible proposal
Eight content sections plus a verified reference list, with page-budget monitoring, comment-driven revision, and DOCX/PDF export. Claims that cannot be supported are corrected or explicitly limited.
Scientific quality gates
What is actually complete — and what is not
Live status · loading
Loading quality gates...
Counts are computed from SQLite, not hardcoded.
Search execution and screening are separate gates.
Corrections are logged in the integrity register.
Provenance
How to read the numbers on this page
Research evidence. References are checked against Crossref/PubMed and, where possible, the full text. Extraction notes are labelled as interpretation until independently re-extracted.
Power. The calculator is a transparent paired-design approximation. It is not a simulation and does not replace a statistician's design review.
Host statements. Only dated host emails or agreements count as commitments. Seeded Q&A entries are workspace notes, not verbatim quotations.
Strategic topic finder
Candidate research topics
Four competing directions evaluated against host capability, published benchmarks and Humboldt criteria. Scores are internal prioritisation heuristics, not validated psychometric or statistical measures.
Active directionTopic A
Selection rule. The active topic drives the proposal draft. Ending endorsement by the host is not the same as a formal host agreement; confirm scope, sample access, genotyping budget and ethics before submission.
Loading candidate topics...
Carlo Maj topic feedback
Host collaborator portal
PD Dr. Carlo Maj review & feasibility triage
A structured review station for laboratory access, sample logistics, genotyping, power and the 18-month work plan. Only items marked Confirmed Fact should be read as host commitments.
Host-evidence rule. The dated host emails in this workspace support blood-sample access, a hosting agreement, a workplace and English as the working language. They do not, by themselves, confirm the final perturbation agent, genotyping budget, radiation-core access or statistical power. Those remain open items.
Commitments checklist
Center for Human Genetics, Marburg
3 confirmed · 1 proposed
1. Blood samples
Host confirmed starting from available blood samples in Marburg so the applicant experiences the experimental workflow firsthand (email, 8 Oct 2026).
Confirmed fact
2. Hosting agreement
Carlo Maj confirmed provision of the official AvH hosting agreement and a designated research workplace (email, 3 Nov 2026).
Confirmed fact
3. English working language
English confirmed as the official working language for the group and proposal (email, 3 Nov 2026).
Confirmed fact
4. Core-facility and genotyping
Radiation-core access, donor genotyping budget and the final perturbation agent are unresolved. Bleomycin is proposed; radiation is a confirmatory option.
Unresolved assumption
Interactive feasibility triage
Update classifications and add notes. Changes save to the canonical database and are audit-logged.
Parameter
Category
Classification
Description
Host guidance / notes
Loading feasibility register...
Official regulations
Alexander von Humboldt Foundation fellowship guidelines
Experienced Researcher track · reference area D201 · call timing and volume cap must be verified on the official portal
Source caution. This workspace records eligibility and timing from the applicant's correspondence. Stipend rates, page limits and the “first 800 applications” cap should be re-checked against the current AvH call document before submission. Do not rely on this page as the formal call text.
1. Eligibility
Doctorate: completed within the last 12 years. Javad completed his PhD in 2019.
Position: Assistant Professor / independent research profile.
Mobility: no extensive stay in Germany in the 18 months before submission.
Formal status: recorded as confirmed via the AvH selection office (22 Sept 2026); retain the written confirmation.
2. Duration & split stays
Total: 6 to 18 months of fellowship funding.
Split visits: up to three stays, each at least three months.
Window: stays completed within 36 months of the first stay.
Stipend: the rate shown in earlier drafts is indicative only; confirm the current AvH rate.
3. Call & proposal specs
Call date: 15 November 2026 (reported).
Volume cap: reported as the first 800 applications; verify this in the official call.
Length: approximately five pages including bibliography (reported).
Structure: state of research, gap, questions, methods, feasibility, host fit, timeline.
Evaluation pillars
Canonical reference
Confirmed foundations & verified benchmarks
Immutable baseline facts, host laboratory profile, and literature benchmarks checked against the source.
Proposed host laboratory
PD Dr. Carlo Maj is Head of Bioinformatics and Research Group Leader in Genetic Epidemiology at the Center for Human Genetics, Philipps-University Marburg. His group works on statistical genetics, exposure-eQTL mapping, and computational functional genomics.
Confirmed: primary blood handling, CD8+ T cell purification, workplace, hosting agreement, English working language.
Not yet confirmed: final perturbation agent, donor genotyping budget, radiation-core access.
Applicant profile
Dr. Javad Alizargar, MD, PhD is an Assistant Professor at the National Taipei University of Nursing and Health Sciences (NTUNHS), Taiwan. He completed his PhD in 2019 and has served as PI on government-funded health research projects.
Humboldt area: D201 – Medical Biometry, Epidemiology, Medical Informatics.
Eligibility: confirmed by the AvH selection office (22 Sept 2026).
Verified published benchmarks
1. DNA-damage response e2QTL benchmark
Bigge et al., BMC Genomics 2024 • 10.1186/s12864-024-11068-6
Primary CD8+ T cells from 461 healthy donors; five carcinogens (MMS, TBOOH, UVC, BPDE, HC); 6 h exposure; 654 regulated genes and 47 e2QTL genes. Not a bleomycin protocol.
2. Monocyte TLR4 response-eQTL benchmark
Kim et al., Nature Communications 2014 • 10.1038/ncomms6236
137 healthy male donors; LPS 90 min; Illumina HumanHT-12 v4. The 432-donor LPS 2 h/24 h study is Fairfax et al., Science 2014 (10.1126/science.1246949).
3. Primary monocyte response-eQTL benchmark
Fairfax et al., Science 2014 • 10.1126/science.1246949
432 European donors; LPS 2 h and 24 h; IFN-gamma 24 h; more than half of cis-eQTLs were stimulus-specific.
4. Macrophage response-eQTL and chromatin benchmark
Alasoo et al., Nature Genetics 2018 • 10.1038/s41588-018-0046-7
Shared chromatin and expression QTLs in macrophages exposed to IFN-gamma, Salmonella and the combination; enhancer priming in immune response.
Literature intelligence
Reproducible multi-API literature discovery
Structured query execution across PubMed, Europe PMC and OpenAlex for all 13 families.
Search execution is not screening. A completed search family means records were retrieved and deduplicated. Inclusion decisions are recorded separately in the library and are currently limited to a small curated set.
Search audit trail
Latest 50 logged executions
Date
Family
Source engine
Query string
Retrieved
New ingested
Loading search logs...
Permanent library
Scientific literature library
DOI-deduplicated records, full-text index, screening status, provenance and export. Title-level duplicates and screening are still being cleaned; counts are live from the database.
Evidence status. Most records are discovered, not included. Inclusion should follow a documented screening decision. Records with corrected metadata are listed in the integrity register.
Loading literature library...
Network view
Citation & thematic network
Nodes and edges are loaded from the database. Click a node to inspect metadata.
Selected publication
Click a node to view its title, authors, citation count and DOI.
Evidence register
Structured evidence extraction
Assay conditions, donor demographics, statistical models and provenance. Verification status is shown on every record.
Extraction is not endorsement. Records marked author interpretation are structured summaries, not verbatim source claims. Re-check the source tables before citing effect sizes or specific assays.
Loading structured evidence...
Adversarial check
Novelty verification
Automated prior-art screens and competing-direction audits. A machine screen is not a novelty verdict.
Manual review required. The current automated screen returns an uncertain verdict and a broad competing-study set. Many returned records are not directly relevant. Do not describe the project as “novel” in the proposal until a human reviewer has screened the competing literature and documented the gap.
Recorded novelty audits
Loading novelty assessments...
Design audit
Statistical power & laboratory feasibility
A transparent paired-design approximation, not a simulation. Use it to expose design limits before the host agreement is finalized.
Headline finding. At N = 70, MAF = 0.20, β = 0.35 and α = 1e-6, the paired interaction model has less than 1% power. The previous 84.2% figure was an error. The proposal now treats WP2 as prespecified candidate-panel discovery plus independent replication, not genome-wide discovery.
Interaction-eQTL power calculator
Loading...
Assumptions
Power curve for the current β and α
Current N80% thresholdOther N
Run the calculator to render the curve.
Bulk RNA-seq Planning estimate: 30M paired-end reads per condition; lower technical dropout; better power per euro for donor-level interaction testing.
Single-cell RNA-seq Resolves sub-lineage heterogeneity, but does not replace donor-level replication. Reserve for a targeted validation subset.
Humboldt research outline
Proposal development workspace
Eight content sections plus a verified reference list. Claims are limited to what the evidence supports; the five-page target is monitored below.
—
Total words
—
Estimated pages
—
Sections + references
—
Page budget
Draft status. The text is a working draft, not a submission-ready proposal. Power claims, host commitments, page limits and bibliographic metadata still require final verification. The export includes a visible draft-status line.
Loading proposal sections...
Audit log
Summary of changes
Timestamped record of section comments, directives and revisions. The integrity audit also writes a revision entry.
Loading revision changelog...
Bilateral discussion
Host–applicant scientific Q&A
Structured questions, answers and consensus records between Dr. Javad Alizargar and PD Dr. Carlo Maj.
Open review mode
Quotation caution. Entries in this workspace are user-authored notes. They are not verbatim quotations from PD Dr. Carlo Maj unless a dated email or signed document is attached. Treat seeded entries as discussion prompts, not host commitments.
Ask a question or post a protocol discussion
Both applicant and host can reply
Scientific inquiries & threaded responses
Loading discussions...
Loading collaboration discussions...
Data infrastructure
Supported scientific APIs & repositories
Authoritative catalog of biomedical literature APIs, genomic repositories and AI synthesis engines integrated into the workspace.
Catalog loaded live
Literature APIs
8
PubMed, Europe PMC, OpenAlex...
Genomic repositories
5
GEO, SRA, BioProject, GTEx, GWAS
Public access
100%
No API key required for public data
AI synthesis
Active
DeepSeek and fallbacks
Loading supported scientific APIs...
Scientific integrity
Corrections, references and provenance
A transparent register of what was corrected, what is verified, and what still needs manual review. This is not a systematic review.
Corrections logged
—
Across power, citations and data
Verified references
—
Crossref / PubMed checked
Critical corrections
—
Must be resolved before submission
Last audit
—
Workspace database
Loading scope statement...
Correction register
Do not quote the “before” text
Loading corrections...
Verified reference register
Bibliographic metadata
Loading references...
What this workspace does not claim
It is not a systematic review or meta-analysis.
It is not a substitute for simulation-based power analysis or a statistician's review.
It does not turn an automated novelty screen into a novelty verdict.
It does not convert workspace notes into host commitments.