page_content stringlengths 10 9.29k | metadata dict |
|---|---|
Joy et. al
2
Figure S2: Performance of BTE‑RAG versus an LLM‑only baseline on the gene‑centric
benchmark using gpt‑4o‑mini.
(A) Overall accuracy as a function of how much of the retrieved context is retained. Bars show accuracy when
only context lines above a given cosine‑similarity percentile are su... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
3
Figure S3: Performance of BTE‑RAG versus an LLM‑only baseline on the gene‑centric
benchmark using gpt‑4o.
(A) Overall accuracy as a function of how much of the retrieved context is retained. Bars show accuracy when
only context lines above a given cosine‑similarity percentile are supplie... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
4
Figure S4: Cosine-similarity profile for the metabolite-centric benchmark using
GPT-4o-mini in LLM-only mode.
Histogram shows the frequency of cosine-similarity scores (bin width = 0.05) between model
answers and ground truth answers across 201 metabolite-related queries when... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
5
Figure S5: Distribution of answer similarities for the metabolite-centric benchmark
using GPT-4o-mini in BTE-RAG mode.
Each panel shows the cosine similarity between model predictions and ground-truth answers when either the full
retrieved context is used (top left) or wh... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
6
Figure S6: Distribution of answer similarities for the metabolite-centric benchmark using
GPT-4o in BTE-RAG mode.
Each panel shows the cosine similarity between model predictions and ground-truth answers
when either the full retrieved context is used (top left) or when cont... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
7
Figure S7: Rank-ordered cosine similarities between model predictions and ground-
truth answers on the metabolite-centric benchmark, across context filtering thresholds.
Each panel displays results from four model–method combinations (GPT-4o-mini-Prompt (LLM-only), GPT-4o-mini ... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
8
Figure S8: Cosine-similarity profile for the drug-centric benchmark using GPT-4o-mini in
LLM-only mode.
Histogram shows the frequency of cosine-similarity scores (bin width = 0.05) between model answers and ground
truth answers across 842 drug-biological process queries whe... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
9
Figure S9: Distribution of answer similarities for the drug-centric benchmark using GPT-
4o-mini in BTE-RAG mode.
Each panel shows the cosine similarity between model predictions and ground-truth answers when either the full
retrieved context is used (top left) or when c... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
10
Figure S10: Distribution of answer similarities for the drug-centric benchmark using
GPT-4o in BTE-RAG mode.
Each panel shows the cosine similarity between model predictions and ground-truth answers
when either the full retrieved context is used (top left) or when cont... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
11
Figure S11: Rank-ordered cosine similarities between model predictions and ground-
truth answers on the drug-centric benchmark, across context filtering thresholds.
Each panel displays results from four model–method combinations (GPT-4o-mini Prompt (LLM-only), GPT-4o-mini
+ B... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
Joy et. al
12
Table S1: System Prompts
Gene-centric benchmark
LLM-only
BTE-RAG
You are an expert biomedical researcher. Please
provide your answer (only gene name) in the
following JSON format for the Question asked:
{
"answer": <correct answer>
}
You are an advanced biomedical researcher. Use ... | {
"author": "janet",
"creationDate": "D:20250731211435-07'00'",
"creationdate": "2025-07-31T21:14:35-07:00",
"creator": "Appligent AppendPDF Pro 5.5",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.08.01.668022v1.full.pdf",
"format": "PDF 1.7",
"keywords": "",
"modDate": "D:20260110171154... |
∗Corresponding author
Email address: suhardivi@hss.edu (Vincentius Jeremy Suhardi)
Evaluating Large Language Models for Gene-to-
Phenotype Mapping: The Critical Role of Full-Text
Database Access
Nicolas Matthew Suhardia, Anastasia Oktarinaa, Julia Retzkya, Damanpreet Dhillond,
Dona Ninane, Mathias P.G. Bostroma,b,... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
2
Introduction
Understanding the relationship between genes and their cellular phenotypes is crucial
in the biomedical field. It provides insights into how genetic perturbations impact cellular
behavior[1]. This knowledge is essential for determining how gene function relates to
cellular phenotypes. The process ... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
3
To address these concerns and evaluate LLM performance in the critical area of gene-
phenotype identification, we assessed transformer-based LLMs—specifically ChatGPT-
4o and Gemini 1.5 Pro. These models were evaluated based on their embedded
biological knowledge, as well as their ability to leverage information ... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
4
prompt format detailed in the LLM Prompts section. Queries were individually executed
via manual entry into browser-based interfaces, ensuring consistent methodology for
each gene-phenotype combination.
Response Collection: Responses generated by the LLMs were systematically collected
by manually transferring ... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
5
Technical Implementation: These specialized models are implemented as closed-
source, publicly available, custom GPTs. Each model utilizes the ChatGPT platform’s
internet search capabilities and custom API integrations as tools. SciSpace and
ScholarAI employ API calls to their respective proprietary databases an... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
6
Queries
Six phenotypes were utilized in this study: bone formation, fibrosis, cartilage formation,
cell proliferation, tendon formation, and ligament formation. These phenotypes were
chosen as representative topics of interest for skeletal and connective tissue
researchers, reflecting both mineralized and soft... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
7
Table 1: List of genes used in this study
Phenotype
Gene Names
Fibrosis
Piezo1, En1, Bglap, Osterix, Wnt1, Wnt3a, Osteopontin, Pth1R,
LRP5, LRP6, RSPO1, RSPO2, RSPO3, Sclerostin, Dkk1, Dkk2,
Piezo2, Spry2, Dkk3, Dkk4, Gli1, Gli2, Gli3, BMP1, BMP2,
Col2a1, Prx1, Sox2, WIF-1, Sp7, Embigin
Bone
Alp, Col1a1, Ru... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
8
LLM Prompts
One-shot prompting approach was utilized to ensure consistent and structured
responses across all models. The prompt design included model-specific instructions to
direct each LLM to use its appropriate database or search capabilities, followed by a
standardized query structure and output format.
... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
9
Figure 1. Standardized prompt template used for querying large language
models. The prompt consists of three main components: (1) Model-specific instructions
(part 1-5) directing each LLM to use its designated database or search capabilities, with
only the relevant instruction provided to each model; (2) Core q... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
10
literature; and (3) Detailed response formatting instructions (part 7) requiring a single-
word answer in square brackets followed by supporting references and impact factor
information. The example output (part 8) demonstrates the expected response format
to ensure consistency across all 1,188 queries per mode... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
11
Statistical Analysis
Logistic regression was performed to analyze the performance of large language
models (LLMs) in providing correct answers, using human responses as the ground
truth. ChatGPT-4o (base model) was treated as the reference model. The p-value was
calculated to test the null hypothesis that t... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
12
1.5 Pro (14.6 vs. 14.0, respectively; p = 0.001, p=0.0001, Figure 2c). Human-provided
references had a statistically significantly higher impact factor than both ChatGPT-4o
and Gemini 1.5 Pro (p = 0.0001 for human vs. ChatGPT-4o, and p = 0.0001 for human
vs. Gemini 1.5 Pro; Figure 2d).
The scientific-focus... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
13
base ChatGPT-4o and human to map genes to phenotype based on relevant peer-
reviewed articles.
ScholarGPT had comparable accuracy compared to ChatGPT-4o (58.6 % for
ScholarGPT vs 59.3 % for ChatGPT-4o, p = 0.064; Figure 4a). Compared to ChatGPT-
4o, ScholarGPT demonstrated significantly lower hallucination rate... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
14
for human; p = 0.0016 for ScholarAI vs. SciSpace; p = 0.001 for human vs. SciSpace
and human vs. ScholarAI; Figure 5d). The phenotype-specific heatmap revealed that
ScholarAI and SciSpace achieved remarkably consistent high performance across all
phenotypes (79.3 - 82.3% for ScholarAI and 71.2 - 80.8% for SciS... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
15
The suboptimal accuracy of these general-purpose LLMs likely results from their
inability to directly access the peer-reviewed articles, restricting their literature retrieval
capabilities and accuracy in synthesizing gene-to-phenotype relationships [24].
Moreover, their inherent design for linguistic rather t... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
16
applicability of these models. Furthermore, the development of standardized protocols
for clinical integration, quality control, and interpretation frameworks is essential for
practical, safe, and effective real-world use.
Despite these promising outcomes, our study has limitations, including the evaluation of
a ... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
17
Figure 2. Accuracy and impact factor of references provided by ChatGPT4o and
GeminiAI. (a) Accuracy of correct gene-to-phenotype mappings by ChatGPT4o and GeminiAI
Each dot represents a single gene-to-phenotype query, with a value of 1 indicating agreement
between the tested LLM and human answers, and 0 indicating... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
18
Figure 4. Accuracy and impact factor of references provided by ScholarGPT. (a) Accuracy
of gene-to-phenotype mappings by ChatGPT4o and ScholarGPT. Each dot represents a single
gene-to-phenotype query, with a value of 1 indicating agreement between the tested LLM and
human answers, and 0 indicating a discrepancy. D... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
19
provided by ChatGPT4o, SciSpace, and ScholarAI. Data are presented as mean ± s.e.m.,
analyzed using the Chi-square test. (d) Impact factor of references provided by human,
ChatGPT4o, SciSpace, and ScholarAI. Each dot represents the impact factor of a single gene-
to-phenotype query. Data are presented as mean ±... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
20
Supplementary Figure 1: Phenotype-specific performance heatmap showing agreement
rates (%) for each model across six different phenotypes. The heatmap reveals consistent
patterns: cartilage formation (lighter yellow regions) poses the greatest challenge across all
models, while fibrosis and cell proliferation ... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
21
Author Contributions
Nicolas Suhardi and Vincentius Suhardi selected the list of genes and phenotypes,
tested the relevance and accuracy of the references provided by the LLMs, conducted
data analysis, and formulated questions. Anastasia Oktarina, Damanpreet Dhillon, and
Dona Ninan recorded the responses gene... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
22
References
[1] A. Dixit, O. Parnas, B. Li, et al., Perturb-seq: Dissecting molecular circuits with
scalable single-cell rna profiling of pooled genetic screens, Cell 167 (7) (2016)
1853–1866.e17. doi:10.1016/j.cell.2016.11.038.
[2] H. Cui, C. Wang, H. Maan, et al., scgpt: toward building a foundation model fo... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
23
[14] Y. Mao, Y.-Y. Lin, N.-K. Y. Wong, et al., Phenotype prediction from single-cell rna-
seq data using attentionbased neural networks, Bioinformatics 40 (2) (2024).
doi:10.1093/bioinformatics/btae067.
[15] M. Elgart, M. O. Goodman, C. Isasi, et al., Correlations between complex human
phenotypes vary by genet... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
24
[25] A. Preisler, Correctness and quality of references generated by ai-based research
assistant tools: The case of scopus ai, elicit, scispace and scite in the field of
business
administration
(2024).
https://www.researchgate.net/publication/383988574_Correctness_
and_Quality_of_References_generated_by_AI... | {
"author": "Administrator",
"creationDate": "D:20250611135555-07'00'",
"creationdate": "2025-06-11T13:55:55-07:00",
"creator": "PScript5.dll Version 5.2.2",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/2025.06.11.659165v1.full.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:2026011... |
D380–D384
Nucleic Acids Research, 2016, Vol. 44, Database issue
Published online 20 November 2015
doi: 10.1093/nar/gkv1277
STITCH 5: augmenting protein–chemical interaction
networks with tissue and affinity data
Damian Szklarczyk1, Alberto Santos2, Christian von Mering1, Lars Juhl Jensen2,
Peer Bork3,4,* and Michael Kuh... | {
"author": "",
"creationDate": "D:20151224224617-05'00'",
"creationdate": "2015-12-24T22:46:17-05:00",
"creator": "XMLPublish v.3.0 Copyright 1991-2015 Aptara Inc.",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/gkv1277.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:20260111010403+... |
Nucleic Acids Research, 2016, Vol. 44, Database issue D381
lows programmatic access, including the ability to disam-
biguate queries, modify all network parameters and gen-
erate images. In order to enable large-scale analysis, which
may not be feasible through web-interface or API, the pre-
computed network and the su... | {
"author": "",
"creationDate": "D:20151224224617-05'00'",
"creationdate": "2015-12-24T22:46:17-05:00",
"creator": "XMLPublish v.3.0 Copyright 1991-2015 Aptara Inc.",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/gkv1277.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:20260111010403+... |
D382 Nucleic Acids Research, 2016, Vol. 44, Database issue
Figure 1. Display of binding affinities. The user interface of STITCH has been updated and the option to scale edge width of protein–chemical interactions
according to binding affinity has been added. The shown network of multiple NSAIDs makes their different b... | {
"author": "",
"creationDate": "D:20151224224617-05'00'",
"creationdate": "2015-12-24T22:46:17-05:00",
"creator": "XMLPublish v.3.0 Copyright 1991-2015 Aptara Inc.",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/gkv1277.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:20260111010403+... |
Nucleic Acids Research, 2016, Vol. 44, Database issue D383
Figure 2. Filtering interaction networks according to tissue expression pat-
terns. (A) The interaction network around diclofenac and PTGS1/2 is
shown without filtering for tissue expression patterns. In this and the fol-
lowing panels, the top five interaction... | {
"author": "",
"creationDate": "D:20151224224617-05'00'",
"creationdate": "2015-12-24T22:46:17-05:00",
"creator": "XMLPublish v.3.0 Copyright 1991-2015 Aptara Inc.",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/gkv1277.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:20260111010403+... |
D384 Nucleic Acids Research, 2016, Vol. 44, Database issue
et al. (2013) The BioGRID interaction database: 2013 update.
Nucleic Acids Res., 41, D816–D823.
14. Kjærulff,S.K., Wich,L., Kringelum,J., Jacobsen,U.P.,
Kouskoumvekaki,I., Audouze,K., Lund,O., Brunak,S., Oprea,T.I.
and Taboureau,O. (2013) ChemProt-2.0: visual n... | {
"author": "",
"creationDate": "D:20151224224617-05'00'",
"creationdate": "2015-12-24T22:46:17-05:00",
"creator": "XMLPublish v.3.0 Copyright 1991-2015 Aptara Inc.",
"file_path": "/home/donbr/ag-discovery/gdelt-kb/data/raw/gkv1277.pdf",
"format": "PDF 1.4",
"keywords": "",
"modDate": "D:20260111010403+... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.