BGPT: Author Review: Yuannyu Zhang

Explore by Goal

Quick Explanation Copied

Yuannyu Zhang — scientific strength (evidence-based, skeptical)

Based on provided publication + metric snapshots, Zhang shows a highly-cited, methods-forward pattern centered on chromatin/epigenetic regulation, gene-regulatory circuitry, and metabolism–epigenome links (e.g., L-2-hydroxyglutarate → MYC epigenetic activation in neurodevelopment; enhancer-targeting CRISPR epigenetic editing; LINE-1 silencing dependencies in myeloid leukemia) .

Long Explanation

Author Review — Yuannyu Zhang

Updated context: provided metrics + provided paper list + one raw-data record (JCI 2026 case study)

1) Evidence snapshot (what we can verify from your input)

OpenAlex identity: “Yuannyu Zhang” (ORCID provided) with works_count = 89, cited_by_count = 2676, h_index = 22 in the OpenAlex record you supplied.
Provided yearly citation pattern (OpenAlex “counts_by_year”) is strongly front-loaded around 2016–2020 with smaller total citations thereafter (notably a year with very low/zero citations in the snapshot).

2) Citation output over time (visual)

Interpretation (skeptical)

Known: your provided OpenAlex snapshot lists works and “cited_by_count” by year.
Uncertainty: the “cited_by_count” by year looks like a snapshot-derived attribution; citation counts can shift with time and indexing delays. Treat year-wise citation values as approximate for trend interpretation, not as immutable ground truth.
Potential blind spot: the snapshot shows very small citations in 2025–2026; this likely reflects recency and indexing, not necessarily lower impact.

3) Thematic evidence from example cited works (with raw-data grounding where available)

A) Metabolism ↔ epigenome ↔ developmental fate (raw-data case)

The JCI-record you provided describes a patient-derived, isogenic correction strategy to link L2HGDH deficiency → L-2HG accumulation → epigenetic changes with reported enrichment at MYC regulatory regions, increased MYC activity, and impaired neuronal differentiation; it also reports rescue/phenocopy logic via MYC depletion and KDM5 inhibition, plus metabolic rescue direction via glutaminase inhibition.

Scientific strength (from what’s explicitly stated): the design emphasizes genetic causality (patient cells + isogenic CRISPR correction) and multi-omic mechanistic triangulation (RNA-seq + ChIP-seq + metabolite quant). This is stronger than purely correlative “omics-only” mechanistic claims.

Limitations / known unknowns (explicit in your record): limited human sample size (two patients), possible iPSC/ex vivo limitations, isotope-tracing limitation regarding L- vs D-2HG discrimination, CRISPR/shRNA off-target risks, and uncertain long-term in vivo translational relevance.

B) Tools for perturbing enhancer function (mechanistic regulatory biology)

Provided OpenAlex top works include enhancer-targeting CRISPR epigenetic editing reported in Nature Communications (2020), positioned as enabling functional analysis of distal cis-regulatory elements (enhancers) by targeted epigenome editing.

Skeptical note: any CRISPR/dCas9 epigenetic editing has potential confounders (local chromatin remodeling vs. broader off-target locus effects; variable editing efficiency; cell-state dependence). Your input does not provide those details for this specific paper, so confidence stays moderate rather than “strong”.

C) Context-specific dependency biology (LINE-1 in myeloid leukemia)

Provided OpenAlex top works include Nature Genetics (2021) reporting a selective dependency on LINE-1 retrotransposon silencing in myeloid leukemia.

Blind spot: your input doesn’t include the specific causality/control experiments (e.g., rescue experiments, off-target discrimination, generality across subtypes). Without those details, the mechanistic claim strength cannot be upgraded beyond moderate.

4) Visual map: reported “what-to-what” mechanisms across the cited examples

A compact directed graph summarizing the kinds of biological causal chains your input explicitly supports.

5) Scientific credit & risks (based on only the evidence provided)

Strengths supported by your dataset

Impact: high citation indicators in the OpenAlex snapshot (2676 cited-by; h-index 22). (Metrics are from your provided record; not DOI-citable.)
Mechanism orientation: the JCI 2026 record explicitly links metabolic cause → chromatin marking → transcription factor activity → differentiation phenotype, and includes multiple causal tests (genetic correction; MYC depletion; KDM5 inhibition; metabolic inhibitor directionality).
Regulatory biology toolmaking: enhancer-targeting CRISPR epigenetic editing is directly represented in your provided top-work list, suggesting the author contributes to experimental strategies for distal regulatory element function.

Key limitations / why citation counts are not the full story

Reproducibility uncertainty: citation counts do not guarantee robustness; they also reflect network effects (field size, relevance, and novelty). Without replication data (not provided here), robustness cannot be directly assessed.
Model-generalization risk: the JCI record itself flags that iPSC-derived and ex vivo mouse models may not fully recapitulate in vivo human brain development.
Enantiomer tracing limitation: the record states that the 13C tracing setup cannot distinguish L- from D-2HG in this setup, which constrains mechanistic inference about enantiomer-specific contributions.
Off-target / causal specificity: CRISPR and shRNA off-target effects remain an always-on threat unless thoroughly quantified; your record lists this as a limitation.

6) Directly checkable resource links (from your provided record)

Raw data access (as stated)

GEO RNA-seq: GSE295010 (as provided)
GEO ChIP-seq: GSE295011 (as provided)
Isotope correction repository: wencgu/nac (as provided)

Optional next step: let BGPT run a targeted “raw-data” analysis loop

Feedback:

Updated: April 10, 2026