Biology subjects

Huang, L. F.

Publications and source records attributed to Huang, L. F..

3 recordsLinked to original sources

bioRxiv2026

Information contracts determine what inverse perturbation benchmarks measure

Inverse perturbation benchmarks rank interventions toward a desired cellular state, but high accuracy can answer a different question from biological prioritization. We present InverseBench, which makes candidate information and endpoint construction explicit, and FSTN, a linear pipeline evaluated under it. With identical inputs, candidates and validation, nested ridge reaches Spearman 0.959 for a transcriptional principal-component proxy but 0.445 for independently measured K562 growth fitness across 7,354 genes. Controlled response-entry availability across seven tasks gives heterogeneous gains, 0.442 for CD4 activation and 0.116 for external fitness. Yet predefined shortlist comparisons show no primary-cutoff penalty in K562 or a melanoma co-culture screen, and secondary cutoffs reverse the K562 ordering. FSTN's nonnegative set selector recovers 0.840 of measured two-gene constituents in three picks, against 0.771 for an unconstrained greedy. The framework separates proxy reconstruction, response-informed prioritization and unseen-intervention ranking: an endpoint-dependent accuracy gap does not by itself mean worse biological decisions.

bioinformatics↗

bioRxiv2022

Cell-type-specific aberrant R-loop accumulation regulates target gene and confers cell-specificity

Aberrant R-loops have been found associated with diverse biological dysfunction, including cancers and neurological disorders. However, there isnt any systematic research to characterize aberrant R-loops at the whole genome level at a large scale. Here, we identified aberrant R-loops, including proliferative and suppressive R-loops of 5 end, body, 3 end respectively for the first time, which are found prevalent and vary across diverse physiological conditions. After that, we proposed a deep neural network-based framework, named Deep R-looper Discriminant to identify aberrant R-loops against housekeeping R-loops. To evaluate the predictive performance of the deep learning framework, we constructed multiple prediction models as benchmarks and it showed our framework achieves robust performance for identifying aberrant R-loops against those normal R-loops. Furthermore, we found the customized Deep R-looper Discriminant was capable of distinguishing between proliferative and suppressive R-loops at 5 end, body, 3 end respectively, outperforming baselines. When inspecting the contribution of epigenetic marks to aberrant R-loops of each class, we inferred landmark epigenetic modifications which play a crucial role in the differentiated formation of those aberrant R-loops, and cell line specificity of epigenetic marks map was found as well. To explore the characteristics of these aberrant R-loops, we depicted the histone landscapes for aberrant R-loops. Finally, we integrated omics and identified target genes regulated directly by aberrant R-loops and found key transcription factors involved in R-loop regulation, which may be implicated in conferring cell-specificity and cancer development and progression.

bioinformatics↗

bioRxiv2020

Three-dimensional mapping identifies distinct vascular niches for myelopoiesis

In contrast to virtually all other tissues in the body the anatomy of differentiation in the bone marrow remains unknown. This is due to the lack of strategies to examine blood cell production in situ, which are required to better understand differentiation, lineage commitment decisions, and to define how spatial organizing cues inform tissue function. Here we developed imaging approaches to map all myeloid cells in whole bones and generated 3D atlases of granulocyte and monocyte/dendritic cell differentiation during homeostasis. We found that myeloid progenitors leave the hematopoietic stem cell niche during differentiation. Granulocyte and monocyte dendritic cell progenitors (MDP) do not interact, instead they localize to different sinusoids where they give rise to clusters of immature cells. MDP cluster with Ly6Clo monocytes and conventional dendritic cells; these localize to a unique subset of colony stimulating factor 1 (CSF1, the major regulator of monopoiesis1) -expressing sinusoids. Csf1 deletion in the vasculature disrupted the MDP clusters and their interaction with sinusoids, leading to reduced MDP numbers and differentiation ability, with subsequent loss of peripheral Ly6Clo monocytes and dendritic cells. These data indicate that there is a specific spatial organization of definitive hematopoiesis and that local cues produced by distinct blood vessels are responsible for this organization. These maps provide a blueprint for in situ analyses of hematopoiesis in blood disorders.

immunology↗