You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: content/contribution-data-visualization.md
+5-27Lines changed: 5 additions & 27 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -29,16 +29,8 @@ To address the fragmentation of the R plotting landscape, we developed a suite o
29
29
***[ggtangle](https://github.com/YuLab-SMU/ggtangle) & [ggflow](https://github.com/YuLab-SMU/ggflow) (Relational & Process Flow):****ggtangle** reimagines network visualization within the tidy framework, while **ggflow** provides a dedicated grammar for flowcharts and transition processes, bridging the gap between static relationships and dynamic workflows.
30
30
***[ggfun](https://cran.r-project.org/package=ggfun) (UX & Utilities):** Provides foundational utilities that enhance the developer and user experience across the entire ecosystem.
<strong>Logic Unification:</strong> By treating plots as first-class algebraic objects, these tools allow for the "compositional" creation of complex figures that were previously impossible or required significant manual effort.
39
-
</td>
40
-
</tr>
41
-
</table>
42
34
43
35
## 2. Specialized Academic Domains: Deep Integration
44
36
@@ -52,16 +44,9 @@ Beyond general-purpose utilities, we have pioneered visualization standards in s
52
44
***Single-Cell & Fine-Scale Omics:**[**ggsc**](https://github.com/YuLab-SMU/ggsc) and [**ivolcano**](https://github.com/YuLab-SMU/ivolcano) address the unique needs of high-resolution data, providing specialized geometries and interactive exploration for single-cell clusters and differential expression.
53
45
***Glycobiology & Complex Carbohydrates:**[**gglycan**](https://github.com/YuLab-SMU/gglycan) introduces a grammar for visualizing complex glycan structures. By supporting standard symbolic nomenclature (e.g., SNFG), it enables researchers to integrate glycomic data with other biological layers, bridging a critical gap in multi-omics synthesis.
<strong>Domain Leadership:</strong> These tools are not mere plotting scripts but are **interpretative frameworks** cited in thousands of studies across <em>Nature</em>, <em>Science</em>, and <em>Cell</em>.
62
-
</td>
63
-
</tr>
64
-
</table>
49
+
65
50
66
51
## 3. Semantic Enrichment & Scientific Communication
67
52
@@ -73,16 +58,9 @@ To bridge the gap between abstract data and human intuition, we developed tools
73
58
***[emojifont](https://cran.r-project.org/package=emojifont), [shadowtext](https://cran.r-project.org/package=shadowtext), & [meme](https://cran.r-project.org/package=meme):** Enhancing semantic storytelling through advanced typography and cultural icons. These tools improve optical clarity through text halos and allow for creative, engaging data interaction, bridging the gap between formal analysis and impactful communication.
74
59
***[hexSticker](https://cran.r-project.org/package=hexSticker):** Revolutionizing how R developers brand their work. **hexSticker** has established the "Hex Logo" as the universal symbol of professional R package development.
<strong>Scientific Communication:</strong> These tools empower researchers to communicate complex data with clarity, impact, and professional polish, fostering broader adoption of open-source science.
Copy file name to clipboardExpand all lines: content/contribution-knowledge-mining.md
+10-57Lines changed: 10 additions & 57 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -24,103 +24,56 @@ Translating massive omics datasets into clinical insights requires more than sim
24
24
25
25
Before knowledge can be mined, it must be quantified. We established the mathematical prerequisites for biological discovery by developing semantic similarity measures that map fuzzy biological concepts onto rigorous mathematical spaces.
<strong>Semantic Quantification:</strong> Through [**GOSemSim**](https://bioconductor.org/packages/GOSemSim), [**DOSE**](https://bioconductor.org/packages/DOSE), and [**meshes**](https://bioconductor.org/packages/meshes), we provided the first comprehensive software suite for measuring functional similarity among gene products across Gene Ontology, Disease Ontology, and MeSH domains. This work, starting from <em>Bioinformatics</em> (2010), enables the prediction of gene functions, disease-gene associations, and drug-repurposing candidates by leveraging the underlying semantic structure of biomedical knowledge.
34
-
</td>
35
-
</tr>
36
-
</table>
29
+
37
30
38
31
39
32
## Pillar 2: The Universal Enrichment Framework — A Paradigm Shift in Theme Discovery
40
33
41
34
A cornerstone of our contribution is the creation of a **universal mechanistic engine** for functional genomics. By establishing [**clusterProfiler**](https://bioconductor.org/packages/clusterProfiler), we shifted the field from static gene-list analysis to dynamic discovery.
<strong>Comparative Theme Discovery:</strong> We pioneered the concept of **Comparative Biological Theme Analysis** (<em>OMICS</em> 2012), allowing researchers to compare functional profiles across complex experimental designs—multiple time points, genotypes, and drug treatments—simultaneously. This paradigm has been integrated into over 40 bioinformatics tools and cited over 40,000 times, becoming the <em>de facto</em> global standard as detailed in our **<em>Nature Protocols</em>** (2024) feature.
50
37
<br><br>
51
38
<strong>Universal Extensibility:</strong> **clusterProfiler** provides a generic platform that breaks the model-organism barrier, supporting thousands of species and any custom functional annotation, from KEGG and GO to Reactome and disease ontologies.
The microbiome represents one of the final frontiers in functional discovery. Our team has developed specialized infrastructures to address the unique challenges of inter-species interactions and metabolic flow in microbial communities.
<strong>Tidy Ecological Mining:</strong> Through [**MicrobiotaProcess**](https://bioconductor.org/packages/MicrobiotaProcess) and [**MicrobiomeProfiler**](https://bioconductor.org/packages/MicrobiomeProfiler), we established a **tidy framework** for microbiome data mining. **MicrobiotaProcess** provides a comprehensive pipeline for ecological analysis (alpha/beta diversity, biomarker discovery), while **MicrobiomeProfiler** enables functional enrichment analysis specifically tailored to the metabolic and genomic context of microbial datasets.
68
-
</td>
69
-
</tr>
70
-
</table>
47
+
71
48
72
49
73
50
## Pillar 4: Deciphering Cistromic Regulation — From Genomic Location to Function
74
51
75
52
Biological function is often encoded in the non-coding genome. We developed the infrastructure to bridge the gap between genomic cis-regulatory positions and biological function.
<strong>Cistromic Annotation:</strong> [**ChIPseeker**](https://bioconductor.org/packages/ChIPseeker) (<em>Bioinformatics</em> 2015) has become a global cornerstone for epigenomic annotation. It provides a comprehensive framework for annotating genomic peaks (ChIP-seq, ATAC-seq, DNase-seq) with biological context and facilitating cross-dataset comparison and mining within the GEO database.
84
-
</td>
85
-
</tr>
86
-
</table>
56
+
87
57
88
58
89
59
## Pillar 5: Knowledge-Driven Discovery in Single-Cell & Spatial Omics
90
60
91
61
As biological data moves toward higher resolution, the integration of prior knowledge becomes essential for interpretability. Our team has pioneered methods to incorporate biological intelligence into single-cell and spatial discovery.
<strong>Biological Latent Spaces:</strong> Through [**MSGNN**](https://github.com/YuLab-SMU/MSGNN), we proposed integrating biological knowledge priors directly into the graph-based community detection process via graph neural networks. This ensures that single-cell clustering outcomes are not only mathematically robust but also align closely with biological interpretations.
97
65
<br><br>
98
66
<strong>Spatial Functional States:</strong> With [**SVP**](https://github.com/YuLab-SMU/SVP), we introduced a novel paradigm for characterizing cell functional states and identifying spatially variable features, enabling the direct mapping of biological functions (e.g., pathways) across tissue architectures.
## Pillar 6: From Abstract Results to Intuitive Insight — Visual Logic
111
71
112
72
A functional discovery is only complete when it can be interpreted. Our team has pioneered the "Visual Logic" of enrichment, transforming abstract statistical tables into rigorous visual narratives.
<strong>Interpretive Synthesis:</strong> Through [**enrichplot**](https://bioconductor.org/packages/enrichplot), we established the visualization standards for functional discovery. By implementing sophisticated geometries (e.g., dot plots, gene-concept networks, upset plots), **enrichplot** enables researchers to visually synthesize complex enrichment results, uncovering the hidden relationships between biological themes and their underlying molecular drivers.
Copy file name to clipboardExpand all lines: content/contribution-tree-data.md
+3-18Lines changed: 3 additions & 18 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -62,12 +62,7 @@ Before our work, tree visualization was largely restricted to topological displa
62
62
63
63
As omics data reached unprecedented scales, our team introduced the **"Data-to-Tree" paradigm** in foundational work published in [*Molecular Biology and Evolution*](https://academic.oup.com/mbe/article-abstract/35/12/3041/5142656) in 2018 (ESI highly cited). This work proposed two comprehensive methods that redefined the integration of heterogeneous data within a unified evolutionary context.
The two methods introduced in 2018 have since evolved from specialized phylogenetic tools into universal visualization standards:
73
68
<ul>
@@ -82,24 +77,14 @@ The two methods introduced in 2018 have since evolved from specialized phylogene
82
77
</ul>
83
78
<br>
84
79
<strong>Programmable Reproducibility:</strong> Our work in <em>iMeta</em> (2022) established the **ggtree** object—a programmable structure that ensures analytical reproducibility by encapsulating trees, data, and visualization directives.
85
-
</td>
86
-
</tr>
87
-
</table>
80
+
88
81
89
82
## Pillar 4: Vertical Generalization — From Phylogeny to General Hierarchy
90
83
91
84
To demonstrate the universal utility of our grammar, we expanded its scope beyond evolutionary biology. By generalizing the framework to encompass all hierarchical structures, we bridged the gap between specialized biological interpretation and general data science.
<strong>Universal Scope:</strong> Through [**ggtreeDendro**](https://bioconductor.org/packages/ggtreeDendro), we extended the phylogenetic grammar to hierarchical clustering and classification/regression trees. This enables the same rigorous data integration methods used in phylogenetics to be applied to any sample-level or feature-level hierarchical relationship (e.g., cell clustering), unifying disparate analytical workflows under a single theoretical umbrella.
0 commit comments