An Overview of Relationship Between Muc13 With Alcohol for Pancreatic Cancer Patients

Introduction

Many researchers are in the search of biomarkers that can predict Pancreatic cancer. The Carbohydrate Antigen or the Cancer An- tigen, CA 19-9 was detected in 1981 as a possible biomarker for resolution of PC. However, this CA 19-9 can have several false positives and hence is not 100% useful. Other subsequent tests may have to be done for confirmation [1]. shows that individu- als who have had type-II diabetes for less than 4 years were at a 50% higher risk of contracting PC than individuals who have had type-II diabetes for greater than 4 years [2]. have concluded that subjects with long standing diabetes have a higher relative risk of PC association [3]. have also found a relationship between diabe- tes and PC.

Many papers debate whether it is EUS or CT that is a better detec- tor of PC [4, 5]. have tried to detect PC via plasma protein profil- ing [6]. have used digital image analysis on differentiating PC and chronic pancreatitis from normal tissue [7]. have used neural net- work in distinguishing between PC from chronic pancreatitis [8]. have used ensemble of decision trees in detecting PCous cells from normal tissue [9]. have used digital image processing and support vector machines in differentiating PCous cells from normal tissue in EUS images [10].

The idea of the above literature study is to suggest that machine learning algorithms have delved into the realms of detection of pancreatic tumors from normal tissue. However, what is worth pointing out is that detection of these PCous cells would not be of much significance because by then, the patients would already have reached a late stage of cancer and would not survive more than a very few years. Pancreatic cancer is one of the cancers that is somewhat difficult to detect at its onset since symptoms do not show and also there are no qualifying biomarkers validated as of date. Hence there is an urgent need for a prediction model for PC to identify and precaution the high-risk group of undergoing fre- quent medical tests.

A huge percentage of pancreatic cancers are being detected at a late stage, giving the patient only a couple of years for survival. It has also been observed from previous works that use of synthetic chemicals, smoking and alcohol history and genetics greatly influ- ence the occurrence of pancreatic cancer [11, 12].

The Dataset

Groupings of the various values in the dataset, after being given a digital value for processing and normalized are shown in table 1.

                                                                          Table 1: Feature values in dataset

Results

We also observe that for mortality status=1 (that is patient sur- vived), the value of features would be as sex=Female, age between 33-39 years, grade and stage =0. This dataset consists of 13 in-put parameters and 2 outputs-survival status and survival no. of months. Following are the 2D plots using t-SNE and Adasyn al- gorithms (considering mortality status as the output variable), as shown in figure 1.

(a)                                                     (b)

Figure 1: Figure showing 2D t-SNE and 2D ADASYN plots

Feature Selection

A total of 15 algorithms were used for the feature selection. Infinite Latent Feature Selection (ILFS), Infinite Feature Selection(InfFS), Eigenvector Centrality Feature Selection(ECFS), Minimum Re- dundancy Maximum Relevance Feature selection(mRMR), Re- lieff, Mutual Information Feature Selection (MutInfFS), Laplacian, Fisher, L2,1-norm Regularized Discriminative Feature Selection for Unsupervised Learning(UDFS), Feature Selection and Kernel Learning for Local Learning-Based Clustering(LLCFS), correla- tion based feature selection(CFS), Unsupervised Feature Selection with Ordinal Locality(UFSOL)[25], Monte Carlo Feature Selec- tion(MCFS), Feature Selection with Adaptive Structure Learn- ing(FSASL) [13-27].

The sum of the priorities defined by these algorithms were summed up to determine the features ranked as per their priority. The results in descending order of priority are: DRINKING, Sex, OverallMCS, NucleusMCS, MembraneMCS, HEPATITIS, Tmn, CytoMCS, Smoking, Grade, Stage, DIABETES, Age. These re- sults were obtained after summing up the ranking given by the different feature selection algorithms and the lowest rank was the feature which has the greatest influence on Pancreatic Cancer, as shown in table 2. Hence drinking definitely influences cause of Pancreatic Cancer and also causes increase in MUC13 toxin in the cells [28–30].

                                                                     Table 2: The Features Considered in The Dataset.

Conclusion

Pancreatic cancer has been found to be directly influenced by smoking history, alcohol abuse, no. of cigarettes smoked in a day, genetics etc. Interestingly, there are certain other less known fea- tures, for example, sex, hepatitis -B, diabetes which are found to also influence causality of cancer in a subtle way.

Acknowledgement

The author of this paper would like to thank Dr Bonny Banerjee and Dr Chrysanthe Preza from the University of Memphis, Ten- nessee for their support and Dr Subhash Chauhan and Dr Sheema Khan from University of Texas, Rio Grand Valley, Texas for shar- ing the dataset and for their guidance in writing the paper.

References

1. Huxley, R., Ansary-Moghaddam, A., Berrington de Gonzalez, A., Barzi, F., & Woodward, M. (2005). Type-II diabetes and pancreatic cancer: a meta-analysis of 36 studies. British jour- nal of cancer, 92(11), 2076-2083.
2. Everhart, J., & Wright, D. (1995). Diabetes mellitus as a risk factor for pancreatic cancer: a meta-analysis. Jama, 273(20), 1605-1609.
3. Ben, Q., Xu, M., Ning, X., Liu, J., Hong, S., Huang, W., ... & Li, Z. (2011). Diabetes mellitus and risk of pancreatic cancer: a meta-analysis of cohort studies. European journal of cancer,47(13), 1928-1937.
4. De Angelis, C., Brizzi, R. F., & Pellicano, R. (2013). Endo- scopic ultrasonography for pancreatic cancer: current and fu- ture perspectives. Journal of Gastrointestinal Oncology, 4(2), 220.
5. DeWitt, J., Devereaux, B. M., Lehman, G. A., Sherman, S., & Imperiale, T. F. (2006). Comparison of endoscopic ultrasound and computed tomography for the preoperative evaluation of pancreatic cancer: a systematic review. Clinical Gastroenter- ology and Hepatology, 4(6), 717-725.
6. Honda, K., Hayashida, Y., Umaki, T., Okusaka, T., Kosuge, T., Kikuchi, S., ... & Yamada, T. (2005). Possible detection of pancreatic cancer by plasma protein profiling. Cancer re- search, 65(22), 10613-10622.
7. Das, A., Nguyen, C. C., Li, F., & Li, B. (2008). Digital image analysis of EUS images accurately differentiates pancreatic cancer from chronic pancreatitis and normal tissue. Gastroin- testinal endoscopy, 67(6), 861-867.
8. SÄ?ftoiu, A., Vilmann, P., Gorunescu, F., Gheonea, D. I., Go- runescu, M., Ciurea, T., ... & Iordache, S. (2008). Neural net- work analysis of dynamic sequences of EUS elastography used for the differential diagnosis of chronic pancreatitis and pancreatic cancer. Gastrointestinal endoscopy, 68(6), 1086-1094.
9. Ge, G., & Wong, G. W. (2008). Classification of premalignant pancreatic cancer mass-spectrometry data using decision tree ensembles. BMC bioinformatics, 9, 1-12.
10. Zhang, M. M., Yang, H., Jin, Z. D., Yu, J. G., Cai, Z. Y., & Li,Z. S. (2010). Differential diagnosis of pancreatic cancer from normal tissue with digital imaging processing and pattern rec- ognition based on a support vector machine of EUS images. Gastrointestinal endoscopy, 72(5), 978-985.
11. Dutta, A. (2023). Using Machine Learning to Identify the Risk Factors of Pancreatic Cancer from the NIH PLCO Dataset.
12. Dutta, A. (2023). Mathematical Modeling of Cancers Using Machine Learning Algorithms. Int J Cancer Res Ther, 8(3), 116-126.
13. Roffo, G., Melzi, S., Castellani, U., & Vinciarelli, A. (2017). Infinite latent feature selection: A probabilistic latent graph- based ranking approach. In Proceedings of the IEEE interna- tional conference on computer vision (pp. 1398-1406).
14. Roffo, G. (2017). Ranking to learn and learning to rank: On the role of ranking in pattern recognition applications. arXiv preprint arXiv:1706.05933.
15. Roffo, G., Melzi, S., & Cristani, M. (2015). Infinite feature selection. In Proceedings of the IEEE international conference on computer vision (pp. 4202-4210).
16. Roffo, G., Melzi, S. (2016). Ranking to learn. In Internation- al Workshop on New Frontiers in Mining Complex Patterns, pages 19–35. Springer.
17. Radovic, M., Ghalwash, M., Filipovic, N., & Obradovic, Z. (2017). Minimum redundancy maximum relevance feature selection approach for temporal gene expression data. BMC bioinformatics, 18(1), 1-14.
18. Urbanowicz, R. J., Meeker, M., La Cava, W., Olson, R. S., & Moore, J. H. (2018). Relief-based feature selection: Introduc- tion and review. Journal of biomedical informatics, 85, 189- 203.
19. Beraha, M., Metelli, A. M., Papini, M., Tirinzoni, A., & Rest- elli, M. (2019, July). Feature selection via mutual informa- tion: New theoretical insights. In 2019 international joint con- ference on neural networks (IJCNN) (pp. 1-9). IEEE.
20. He, X., Cai, D., & Niyogi, P. (2005). Laplacian score for fea- ture selection. Advances in neural information processing systems, 18.
21. Gu, Q., Li, Z., & Han, J. (2012). Generalized fisher score for feature selection. arXiv preprint arXiv:1202.3725.
22. Yang, Y., Shen, H. T., Ma, Z., Huang, Z., & Zhou, X. (2011, December). â?? 2, 1-norm regularized discriminative feature se- lection for unsupervised learning. In IJCAI international joint conference on artificial intelligence.
23. Zeng, H., & Cheung, Y. M. (2010). Feature selection and ker- nel learning for local learning-based clustering. IEEE trans- actions on pattern analysis and machine intelligence, 33(8), 1532-1547.
24. Hall, M. A. (1999). Correlation-based feature selection for machine learning (Doctoral dissertation, The University of Waikato).
25. Guo, J., & Zhu, W. (2018, April). Dependence guided unsu- pervised feature selection. In Proceedings of the AAAI con- ference on artificial intelligence (Vol. 32, No. 1).
26. DramiÃ…?ski, M., Rada-Iglesias, A., Enroth, S., Wadelius, C., Koronacki, J., & Komorowski, J. (2008). Monte Carlo feature selection for supervised classification. Bioinformatics, 24(1), 110-117.
27. Du, L., & Shen, Y. D. (2015, August). Unsupervised feature selection with adaptive structure learning. In Proceedings of the 21th ACM SIGKDD international conference on knowl- edge discovery and data mining (pp. 209-218).
28. Yallapu, M. M., Ebeling, M. C., Khan, S., Sundram, V., Chauhan, N., Gupta, B. K., ... & Chauhan, S. C. (2013). Novel cur- cumin-loaded magnetic nanoparticles for pancreatic cancer treatment. Molecular cancer therapeutics, 12(8), 1471-1480.
29. Khan, S., Ebeling, M. C., Chauhan, N., Thompson, P. A.,Gara, R. K., Ganju, A., ... & Chauhan, S. C. (2015). Ormel- oxifene suppresses desmoplasia and enhances sensitivity of gemcitabine in pancreatic cancer. Cancer research, 75(11), 2292-2304.
30. Khan, S., Zafar, N., Khan, S. S., Setua, S., Behrman, S. W., Stiles, Z. E., ... & Chauhan, S. C. (2018). Clinical significance of MUC13 in pancreatic ductal adenocarcinoma. Hpb, 20(6), 563-572.