Predicting Drug Responses by Propagating Interactions through Text-Enhanced Drug-Gene Networks

Shiyin Wang

arxiv: 1906.08089 · v1 · pith:PL467JV4new · submitted 2019-06-19 · 💻 cs.SI

Predicting Drug Responses by Propagating Interactions through Text-Enhanced Drug-Gene Networks

Shiyin Wang This is my paper

Pith reviewed 2026-05-25 19:59 UTC · model grok-4.3

classification 💻 cs.SI

keywords drug responsedrug-gene networktext miningexplainable predictioncell linesinteraction propagationpersonalized medicine

0 comments

The pith

A drug-gene network built from research article patterns predicts drug sensitivity from gene records at 94.74% accuracy.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper shows how to build a drug-gene interaction network by mining patterns from biological research articles and combining them with categorical data. Cell line experimental records are used to estimate edge embeddings in the network. Predictions of drug response are made by propagating interactions through this network, providing white-box explanations based on gene records. The model reaches 94.74% accuracy in distinguishing drug sensitive from resistant cases. Readers would care if this approach allows using published knowledge to improve personalized drug selection without black-box models.

Core claim

The central discovery is that a text-enhanced drug-gene network, constructed from article-mined interactions and calibrated with cell line data, supports accurate and explainable prediction of drug responses via interaction propagation.

What carries the argument

The text-enhanced drug-gene interaction network with estimated edge embeddings from cell line records, which carries the propagation of interactions for response prediction.

If this is right

Predictions of drug sensitivity become directly traceable to specific gene-drug interactions in the network.
The model integrates literature-derived knowledge with experimental data for better performance.
White-box nature allows users to understand why a particular drug is predicted to be effective or not for a gene profile.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the network captures general biological mechanisms, it could be tested on patient-derived data beyond cell lines.
Similar text-enhanced networks might apply to other prediction tasks like disease-gene associations.

Load-bearing premise

The assumption that article-mined patterns and cell line records together form a network sufficient to predict real-world drug responses accurately.

What would settle it

A drop in prediction accuracy below 94.74% when the model is evaluated on independent clinical patient data with known drug responses.

Figures

Figures reproduced from arXiv: 1906.08089 by Shiyin Wang.

**Figure 2.** Figure 2: The visualization of all mined interactions among genes and drugs. The label of the genes are their Entrez ID. The [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 4.** Figure 4: 28 entities are contained in cell line records, which are colored with red. We extracted them and their neighbors [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

read the original abstract

Personalized drug response has received public awareness in recent years. How to combine gene test result and drug sensitivity records is regarded as essential in the real-world implementation. Research articles are good sources to train machine predicting, inference, reasoning, etc. In this project, we combine the patterns mined from biological research articles and categorical data to construct a drug-gene interaction network. Then we use the cell line experimental records on gene and drug sensitivity to estimate the edge embeddings in the network. Our model provides white-box explainable predictions of drug response based on gene records, which achieves 94.74% accuracy in binary drug sensitivity prediction task.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper applies text-mined drug-gene edges to a network and reports 94.74% binary accuracy on cell-line sensitivity data, but the validation details are missing and the real-world claim rests on an untested cell-line to patient jump.

read the letter

The paper constructs a drug-gene interaction network from patterns in research articles, then estimates edge embeddings using cell-line drug sensitivity records to produce predictions. It presents this as a white-box model that reaches 94.74% accuracy on a binary sensitivity task and ties the work to personalized medicine needs. The integration of literature text with experimental records is a straightforward step that some groups in bioinformatics might find useful for building explainable graphs. The attempt to keep the model interpretable rather than purely black-box is also a reasonable direction. Beyond that, the technical steps follow patterns already common in network propagation work on drug data, with no indication of a distinct new mechanism or result. The reported accuracy comes with no description of data splits, baseline comparisons, error bars, or checks against overfitting. The embeddings are derived from the same cell-line records used for evaluation, which raises a direct circularity issue that the abstract does not resolve. More critically, the abstract frames the output as relevant to real-world drug response and gene-test implementation, yet all numbers come from in-vitro cell lines. Those records omit tumor microenvironment, pharmacokinetics, and patient variation, so the transfer to clinical outcomes is an assumption without supporting evidence in the description. This gap is load-bearing for the main claim. The work would mainly interest a small set of researchers already running similar drug-repurposing pipelines who want to see one more text-augmented variant. It does not supply enough controls or external validation to justify time from a broader audience or a serious referee. I would not bring it to a reading group or cite it.

Referee Report

2 major / 1 minor

Summary. The paper constructs a drug-gene interaction network by mining patterns from biological research articles combined with categorical data, estimates edge embeddings using cell line drug sensitivity records, and claims to deliver white-box explainable predictions of drug response from gene records, achieving 94.74% accuracy on a binary drug sensitivity prediction task.

Significance. If the result holds under proper validation, the approach could offer an interpretable method for integrating literature-derived networks with experimental records to support drug response prediction, with potential value in network-based modeling within bioinformatics and personalized medicine applications.

major comments (2)

[Abstract] Abstract: The reported 94.74% accuracy in binary drug sensitivity prediction is obtained by estimating edge embeddings directly from the same cell line experimental records used for evaluation; without details on held-out testing, independent benchmarks, train/test splits, or controls for overfitting, the performance figure cannot be assessed for generalization.
[Abstract] Abstract: The central claim asserts predictions of real-world drug responses based on gene records, yet all data and evaluation derive from cell-line records that omit tumor microenvironment, pharmacokinetics, and patient heterogeneity; this untested extrapolation from in-vitro embeddings to clinical outcomes is load-bearing for the stated applicability.

minor comments (1)

[Abstract] The manuscript provides no information on baselines, error bars, or comparison methods for the accuracy claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed comments on validation procedures and the scope of applicability. We address each point below and will revise the manuscript accordingly to improve clarity and accuracy.

read point-by-point responses

Referee: [Abstract] Abstract: The reported 94.74% accuracy in binary drug sensitivity prediction is obtained by estimating edge embeddings directly from the same cell line experimental records used for evaluation; without details on held-out testing, independent benchmarks, train/test splits, or controls for overfitting, the performance figure cannot be assessed for generalization.

Authors: We agree that the abstract provides insufficient information on the validation setup. The edge embeddings were derived from cell-line records, and the reported accuracy reflects performance on those records. The manuscript will be revised to explicitly describe the data partitioning (including any train/test splits or cross-validation), report additional metrics, and note the absence of fully independent held-out benchmarks if none were used. This will allow proper assessment of generalization. revision: yes
Referee: [Abstract] Abstract: The central claim asserts predictions of real-world drug responses based on gene records, yet all data and evaluation derive from cell-line records that omit tumor microenvironment, pharmacokinetics, and patient heterogeneity; this untested extrapolation from in-vitro embeddings to clinical outcomes is load-bearing for the stated applicability.

Authors: The referee is correct that the abstract and introductory framing imply broader clinical relevance than the experiments support. All results are based on cell-line data. We will revise the abstract, introduction, and conclusions to restrict claims to in-vitro drug sensitivity prediction in cell lines and to explicitly list the unaddressed factors (tumor microenvironment, pharmacokinetics, patient heterogeneity) as limitations on translation to real-world patient outcomes. revision: yes

Circularity Check

1 steps flagged

Edge embeddings estimated from cell-line records; reported accuracy reduces to in-sample fit

specific steps

fitted input called prediction [Abstract]
"we use the cell line experimental records on gene and drug sensitivity to estimate the edge embeddings in the network. Our model provides white-box explainable predictions of drug response based on gene records, which achieves 94.74% accuracy in binary drug sensitivity prediction task."

Edge embeddings are fitted to the identical cell-line sensitivity records that supply the binary labels for the reported accuracy. Without an independent test partition or external validation set stated, the 94.74% figure is the in-sample reconstruction error of the fitted embeddings rather than a genuine out-of-sample prediction.

full rationale

The paper constructs the network from text-mined articles, then estimates edge embeddings directly from cell-line drug-sensitivity records and reports 94.74% accuracy on the binary prediction task. No description of held-out test sets, cross-validation splits, or external benchmarks is provided in the abstract or claimed derivation, so the accuracy is statistically forced by the fitting step itself. This matches the fitted-input-called-prediction pattern and matches the reader's 6.0 assessment. The real-world clinical claim is an untested extrapolation but is not itself a circularity reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no explicit free parameters, axioms, or invented entities can be identified from the provided text.

pith-pipeline@v0.9.0 · 5624 in / 1043 out tokens · 21622 ms · 2026-05-25T19:59:27.636811+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages · 7 internal anchors

[1]

Meng Jiang 0001, Jingbo Shang, Taylor Cassidy, Xiang Ren, Lance M Kaplan, Timothy P Hanratty, and Jiawei Han 0001. 2017. MetaPAD - Meta Pattern Discovery from Massive Text Corpora. CoRR cs.CL (2017)

work page 2017
[2]

Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural ma- chine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473 (2014)

work page internal anchor Pith review Pith/arXiv arXiv 2014
[3]

Jordi Barretina, Giordano Caponigro, Nicolas Stransky, Kavitha Venkatesan, Adam A Margolin, Sungjoon Kim, Christopher J Wilson, Joseph Lehár, Gre- gory V Kryukov, Dmitriy Sonkin, Anupama Reddy, Manway Liu, Lauren Murray, Michael F Berger, John E Monahan, Paula Morais, Jodi Meltzer, Adam Korejwa, Judit Jané-Valbuena, Felipa A Mapa, Joseph Thibault, Eva Bri...

work page 2012
[4]

A. Basu, N. E. Bodycombe, J. H. Cheah, E. V. Price, K. Liu, G. I. Schaefer, R. Y. Ebright, M. L. Stewart, D. Ito, S. Wang, A. L. Bracha, T. Liefeld, M. Wawer, J. C. Gilbert, A. J. Wilson, N. Stransky, G. V. Kryukov, V. Dancik, J. Barretina, L. A. Garraway, C. S. Hon, B. Munoz, J. A. Bittker, B. R. Stockwell, D. Khabele, A. M. Stern, P. A. Clemons, A. F. S...

work page 2013
[5]

Kyunghyun Cho, Bart Van Merriënboer, Caglar Gulcehre, Dzmitry Bahdanau, Fethi Bougares, Holger Schwenk, and Yoshua Bengio. 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078 (2014)

work page internal anchor Pith review Pith/arXiv arXiv 2014
[6]

Allan Peter Davis, Cynthia J Grondin, Robin J Johnson, Daniela Sciaky, Benjamin L King, Roy McMorran, Jolene Wiegers, Thomas C Wiegers, and Carolyn J Mat- tingly. 2017. The Comparative Toxicogenomics Database: update 2017. Nucleic Acids Research 45, D1 (Jan. 2017), D972–D978

work page 2017
[7]

Jacob Devlin, Hao Cheng, Hao Fang, Saurabh Gupta, Li Deng, Xiaodong He, Geof- frey Zweig, and Margaret Mitchell. 2015. Language models for image captioning: The quirks and what works. arXiv preprint arXiv:1505.01809 (2015)

work page internal anchor Pith review Pith/arXiv arXiv 2015
[8]

Jürgen Drews. 2000. Drug discovery: a historical perspective. Science 287, 5460 (2000), 1960–1964

work page 2000
[9]

Zachary C Lipton, John Berkowitz, and Charles Elkan. 2015. A critical review of recurrent neural networks for sequence learning. arXiv preprint arXiv:1506.00019 (2015)

work page internal anchor Pith review Pith/arXiv arXiv 2015
[10]

Minh-Thang Luong, Hieu Pham, and Christopher D Manning. 2015. Effec- tive approaches to attention-based neural machine translation. arXiv preprint arXiv:1508.04025 (2015)

work page internal anchor Pith review Pith/arXiv arXiv 2015
[11]

Jörg Menche, Emre Guney, Amitabh Sharma, Patrick J Branigan, Matthew J Loza, Frédéric Baribaud, Radu Dobrin, and Albert-László Barabási. 2017. Integrating personalized gene expression profiles into predictive disease-associated gene pools. npj Systems Biology and Applications 3, 1 (March 2017), 10

work page 2017
[12]

Reza Mirnezami, Jeremy Nicholson, and Ara Darzi. 2012. Preparing for precision medicine. New England Journal of Medicine 366, 6 (2012), 489–491

work page 2012
[13]

M. G. Rees, B. Seashore-Ludlow, J. H. Cheah, D. J. Adams, E. V. Price, S. Gill, S. Javaid, M. E. Coletti, V. L. Jones, N. E. Bodycombe, C. K. Soule, B. Alexander, A. Li, P. Montgomery, J. D. Kotz, C. S. Hon, B. Munoz, T. Liefeld, V. Dan?ik, D. A. Haber, C. B. Clish, J. A. Bittker, M. Palmer, B. K. Wagner, P. A. Clemons, A. F. Shamji, and S. L. Schreiber. ...

work page 2016
[14]

Steven J Rennie, Etienne Marcheret, Youssef Mroueh, Jarret Ross, and Vaibhava Goel. 2017. Self-critical sequence training for image captioning. In CVPR, Vol. 1. 3

work page 2017
[15]

Brinton Seashore-Ludlow, Matthew G Rees, Jaime H Cheah, Murat Cokol, Ed- mund V Price, Matthew E Coletti, Victor Jones, Nicole E Bodycombe, Christian K Soule, Joshua Gould, et al. 2015. Harnessing connectivity in a large-scale small- molecule sensitivity dataset. Cancer discovery 5, 11 (2015), 1210–1223

work page 2015
[17]

Automated Phrase Mining from Massive Text Corpora

Automated Phrase Mining from Massive Text Corpora. arXiv.org (Feb. 2017). arXiv:cs.CL/1702.04457v2

work page internal anchor Pith review Pith/arXiv arXiv 2017
[18]

Jingbo Shang, Jialu Liu, Meng Jiang, Xiang Ren, Clare R Voss, and Jiawei Han

work page
[19]

IEEE Transactions on Knowledge and Data Engineering 30, 10 (2018), 1825–1837

Automated phrase mining from massive text corpora. IEEE Transactions on Knowledge and Data Engineering 30, 10 (2018), 1825–1837

work page 2018
[20]

Jingbo Shang, Meng Qu, Jialu Liu, Lance M Kaplan, Jiawei Han, and Jian Peng

work page
[21]

Meta-Path Guided Embedding for Similarity Search in Large-Scale Heterogeneous Information Networks

Meta-Path Guided Embedding for Similarity Search in Large-Scale Hetero- geneous Information Networks. arXiv.org (Oct. 2016). arXiv:cs.SI/1610.09769v1

work page internal anchor Pith review Pith/arXiv arXiv 2016
[22]

Dibakar Sigdel, Vincent Kyi, Aiden Zhang, Shaun P Setty, David A Liem, Yu Shi, Xuan Wang, Jiaming Shen, Wei Wang, JiaWei Han, et al . 2019. Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications. JoVE (Journal of Visualized Experiments) 144 (2019), e59108

work page 2019
[23]

Xuan Wang, Yu Zhang, Qi Li, Yinyin Chen, and Jiawei Han. 2018. Open in- formation extraction with meta-pattern discovery in biomedical literature. In Proceedings of the 2018 ACM International Conference on Bioinformatics, Compu- tational Biology, and Health Informatics . ACM, 291–300

work page 2018
[24]

Xuan Wang, Yu Zhang, Qi Li, Yinyin Chen, and Jiawei Han. 2018. Open Informa- tion Extraction with Meta-pattern Discovery in Biomedical Literature . ACM, New York, New York, USA

work page 2018
[25]

Xuan Wang, Yu Zhang, Qi Li, Cathy H Wu, and Jiawei Han. 2018. PENNER: Pattern-enhanced Nested Named Entity Recognition in Biomedical Literature. In 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) . IEEE, 540–547

work page 2018
[26]

Chih-Hsuan Wei, Hung-Yu Kao, and Zhiyong Lu. 2013. PubTator: a web-based text mining tool for assisting biocuration. Nucleic acids research 41, W1 (2013), W518–W522

work page 2013
[27]

Haixiu Yang, Yunpeng Zhang, Jiasheng Wang, Tan Wu, Siyao Liu, Yanjun Xu, and Desi Shang. 2018. Global view of a drug-sensitivity gene network. Oncotarget 9, 3 (Jan. 2018), 3254–3266

work page 2018
[28]

Wanjuan Yang, Jorge Soares, Patricia Greninger, Elena J Edelman, Howard Light- foot, Simon Forbes, Nidhi Bindal, Dave Beare, James A Smith, I Richard Thompson, Sridhar Ramaswamy, P Andrew Futreal, Daniel A Haber, Michael R Stratton, Cyril Benes, Ultan McDermott, and Mathew J Garnett. 2013. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for ther...

work page 2013
[29]

Quanzeng You, Hailin Jin, Zhaowen Wang, Chen Fang, and Jiebo Luo. 2016. Image captioning with semantic attention. In Proceedings of the IEEE conference on computer vision and pattern recognition . 4651–4659. 5

work page 2016

[1] [1]

Meng Jiang 0001, Jingbo Shang, Taylor Cassidy, Xiang Ren, Lance M Kaplan, Timothy P Hanratty, and Jiawei Han 0001. 2017. MetaPAD - Meta Pattern Discovery from Massive Text Corpora. CoRR cs.CL (2017)

work page 2017

[2] [2]

Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural ma- chine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473 (2014)

work page internal anchor Pith review Pith/arXiv arXiv 2014

[3] [3]

Jordi Barretina, Giordano Caponigro, Nicolas Stransky, Kavitha Venkatesan, Adam A Margolin, Sungjoon Kim, Christopher J Wilson, Joseph Lehár, Gre- gory V Kryukov, Dmitriy Sonkin, Anupama Reddy, Manway Liu, Lauren Murray, Michael F Berger, John E Monahan, Paula Morais, Jodi Meltzer, Adam Korejwa, Judit Jané-Valbuena, Felipa A Mapa, Joseph Thibault, Eva Bri...

work page 2012

[4] [4]

A. Basu, N. E. Bodycombe, J. H. Cheah, E. V. Price, K. Liu, G. I. Schaefer, R. Y. Ebright, M. L. Stewart, D. Ito, S. Wang, A. L. Bracha, T. Liefeld, M. Wawer, J. C. Gilbert, A. J. Wilson, N. Stransky, G. V. Kryukov, V. Dancik, J. Barretina, L. A. Garraway, C. S. Hon, B. Munoz, J. A. Bittker, B. R. Stockwell, D. Khabele, A. M. Stern, P. A. Clemons, A. F. S...

work page 2013

[5] [5]

Kyunghyun Cho, Bart Van Merriënboer, Caglar Gulcehre, Dzmitry Bahdanau, Fethi Bougares, Holger Schwenk, and Yoshua Bengio. 2014. Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078 (2014)

work page internal anchor Pith review Pith/arXiv arXiv 2014

[6] [6]

Allan Peter Davis, Cynthia J Grondin, Robin J Johnson, Daniela Sciaky, Benjamin L King, Roy McMorran, Jolene Wiegers, Thomas C Wiegers, and Carolyn J Mat- tingly. 2017. The Comparative Toxicogenomics Database: update 2017. Nucleic Acids Research 45, D1 (Jan. 2017), D972–D978

work page 2017

[7] [7]

Jacob Devlin, Hao Cheng, Hao Fang, Saurabh Gupta, Li Deng, Xiaodong He, Geof- frey Zweig, and Margaret Mitchell. 2015. Language models for image captioning: The quirks and what works. arXiv preprint arXiv:1505.01809 (2015)

work page internal anchor Pith review Pith/arXiv arXiv 2015

[8] [8]

Jürgen Drews. 2000. Drug discovery: a historical perspective. Science 287, 5460 (2000), 1960–1964

work page 2000

[9] [9]

Zachary C Lipton, John Berkowitz, and Charles Elkan. 2015. A critical review of recurrent neural networks for sequence learning. arXiv preprint arXiv:1506.00019 (2015)

work page internal anchor Pith review Pith/arXiv arXiv 2015

[10] [10]

Minh-Thang Luong, Hieu Pham, and Christopher D Manning. 2015. Effec- tive approaches to attention-based neural machine translation. arXiv preprint arXiv:1508.04025 (2015)

work page internal anchor Pith review Pith/arXiv arXiv 2015

[11] [11]

Jörg Menche, Emre Guney, Amitabh Sharma, Patrick J Branigan, Matthew J Loza, Frédéric Baribaud, Radu Dobrin, and Albert-László Barabási. 2017. Integrating personalized gene expression profiles into predictive disease-associated gene pools. npj Systems Biology and Applications 3, 1 (March 2017), 10

work page 2017

[12] [12]

Reza Mirnezami, Jeremy Nicholson, and Ara Darzi. 2012. Preparing for precision medicine. New England Journal of Medicine 366, 6 (2012), 489–491

work page 2012

[13] [13]

M. G. Rees, B. Seashore-Ludlow, J. H. Cheah, D. J. Adams, E. V. Price, S. Gill, S. Javaid, M. E. Coletti, V. L. Jones, N. E. Bodycombe, C. K. Soule, B. Alexander, A. Li, P. Montgomery, J. D. Kotz, C. S. Hon, B. Munoz, T. Liefeld, V. Dan?ik, D. A. Haber, C. B. Clish, J. A. Bittker, M. Palmer, B. K. Wagner, P. A. Clemons, A. F. Shamji, and S. L. Schreiber. ...

work page 2016

[14] [14]

Steven J Rennie, Etienne Marcheret, Youssef Mroueh, Jarret Ross, and Vaibhava Goel. 2017. Self-critical sequence training for image captioning. In CVPR, Vol. 1. 3

work page 2017

[15] [15]

Brinton Seashore-Ludlow, Matthew G Rees, Jaime H Cheah, Murat Cokol, Ed- mund V Price, Matthew E Coletti, Victor Jones, Nicole E Bodycombe, Christian K Soule, Joshua Gould, et al. 2015. Harnessing connectivity in a large-scale small- molecule sensitivity dataset. Cancer discovery 5, 11 (2015), 1210–1223

work page 2015

[16] [17]

Automated Phrase Mining from Massive Text Corpora

Automated Phrase Mining from Massive Text Corpora. arXiv.org (Feb. 2017). arXiv:cs.CL/1702.04457v2

work page internal anchor Pith review Pith/arXiv arXiv 2017

[17] [18]

Jingbo Shang, Jialu Liu, Meng Jiang, Xiang Ren, Clare R Voss, and Jiawei Han

work page

[18] [19]

IEEE Transactions on Knowledge and Data Engineering 30, 10 (2018), 1825–1837

Automated phrase mining from massive text corpora. IEEE Transactions on Knowledge and Data Engineering 30, 10 (2018), 1825–1837

work page 2018

[19] [20]

Jingbo Shang, Meng Qu, Jialu Liu, Lance M Kaplan, Jiawei Han, and Jian Peng

work page

[20] [21]

Meta-Path Guided Embedding for Similarity Search in Large-Scale Heterogeneous Information Networks

Meta-Path Guided Embedding for Similarity Search in Large-Scale Hetero- geneous Information Networks. arXiv.org (Oct. 2016). arXiv:cs.SI/1610.09769v1

work page internal anchor Pith review Pith/arXiv arXiv 2016

[21] [22]

Dibakar Sigdel, Vincent Kyi, Aiden Zhang, Shaun P Setty, David A Liem, Yu Shi, Xuan Wang, Jiaming Shen, Wei Wang, JiaWei Han, et al . 2019. Cloud-Based Phrase Mining and Analysis of User-Defined Phrase-Category Association in Biomedical Publications. JoVE (Journal of Visualized Experiments) 144 (2019), e59108

work page 2019

[22] [23]

Xuan Wang, Yu Zhang, Qi Li, Yinyin Chen, and Jiawei Han. 2018. Open in- formation extraction with meta-pattern discovery in biomedical literature. In Proceedings of the 2018 ACM International Conference on Bioinformatics, Compu- tational Biology, and Health Informatics . ACM, 291–300

work page 2018

[23] [24]

Xuan Wang, Yu Zhang, Qi Li, Yinyin Chen, and Jiawei Han. 2018. Open Informa- tion Extraction with Meta-pattern Discovery in Biomedical Literature . ACM, New York, New York, USA

work page 2018

[24] [25]

Xuan Wang, Yu Zhang, Qi Li, Cathy H Wu, and Jiawei Han. 2018. PENNER: Pattern-enhanced Nested Named Entity Recognition in Biomedical Literature. In 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) . IEEE, 540–547

work page 2018

[25] [26]

Chih-Hsuan Wei, Hung-Yu Kao, and Zhiyong Lu. 2013. PubTator: a web-based text mining tool for assisting biocuration. Nucleic acids research 41, W1 (2013), W518–W522

work page 2013

[26] [27]

Haixiu Yang, Yunpeng Zhang, Jiasheng Wang, Tan Wu, Siyao Liu, Yanjun Xu, and Desi Shang. 2018. Global view of a drug-sensitivity gene network. Oncotarget 9, 3 (Jan. 2018), 3254–3266

work page 2018

[27] [28]

Wanjuan Yang, Jorge Soares, Patricia Greninger, Elena J Edelman, Howard Light- foot, Simon Forbes, Nidhi Bindal, Dave Beare, James A Smith, I Richard Thompson, Sridhar Ramaswamy, P Andrew Futreal, Daniel A Haber, Michael R Stratton, Cyril Benes, Ultan McDermott, and Mathew J Garnett. 2013. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for ther...

work page 2013

[28] [29]

Quanzeng You, Hailin Jin, Zhaowen Wang, Chen Fang, and Jiebo Luo. 2016. Image captioning with semantic attention. In Proceedings of the IEEE conference on computer vision and pattern recognition . 4651–4659. 5

work page 2016