Al Imtiaz
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Al Imtiaz has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Imran Chowdhury
Northern University Bangladesh (NUB)
5 shared publications - Khalid Mahbub Khan
Dept. of Computer Science and Engineering, East West University
4 shared publications - Taslim Ahmed
University of Information Technology and Sciences (UITS)
2 shared publications A. S. M. Shafi
University of Information Technology & Sciences (UITS)
1 shared publicationAbu Z. M. Dayem Ullah
1 shared publication- Jalal Uddin
City University Bangladesh
1 shared publication
Latest Publications
- Open Access
- 26 October 2026
- Published by Elsevier in SSRN Electronic Journal
Clinico-Genomic Integration with Functional-Impact Encoding of Somatic Mutations Identifies HMCN1 as a Candidate Prognostic Factor in Pancreatic Cancer
Pancreatic cancer remains one of the most lethal malignancies, with most patients diagnosed at advanced, inoperable stages. We assessed whether somatic mutations add prognostic value to clinical variables, and whether encoding mutations by their functional impact rather than by binary presence changes which genes are identified as prognostic.Methods: Clinical and somatic mutation data from The Cancer Genome Atlas (TCGA) and the International Cancer Genome Consortium (ICGC) were harmonised into a dataset of 796 patients (discovery: n=597; validation: n=199). Of 43,871 genes, 24 with mutation prevalence ≥5% were analysed. Each gene was encoded either as binary mutation presence (present/absent) or as mutation functional impact (high/low; scored with SIFT, PolyPhen and VEP). Genes were screened with log-rank tests and Benjamini–Hochberg false-discovery-rate correction. Survival was modelled with multivariate Cox proportional-hazards regression and twelve machine-learning (ML) classifiers at 12, 24 and 36 months, using horizon-specific outcomes that excluded patients censored before each horizon. Discrimination on the held-out validation cohort was measured by AUROC (area under the receiver operating characteristic curve) with bootstrap 95% confidence intervals.Results: TP53 was significant after Benjamini–Hochberg correction (q=0.005) and in multivariate analysis. HMCN1 was significant in the multivariate model only under functional-impact encoding (p=0.022; q=0.056) and was not associated with survival under presence encoding. On the held-out cohort, the Cox model achieved an average AUROC of 0.603 and the best ML classifier (AdaBoost) 0.549 (0.638 on cross-validation); both were at chance level at 12 months. Adding mutations to clinical variables did not improve held-out discrimination for either Cox or ML models.Somatic mutation data alone were insufficient for reliable survival prediction in pancreatic cancer and added no discrimination beyond clinical variables. Functional-impact encoding identified HMCN1 as a candidate prognostic factor that requires validation in independent cohorts. Code and data are available on Zenodo.
- 26 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessClinico-Genomic Integration with Functional-Impact Encoding of Somatic Mutations Identifies HMCN1 as a Candidate Prognostic Factor in Pancreatic Cancer
Pancreatic cancer remains one of the most lethal malignancies, with most patients diagnosed at advanced, inoperable stages. We assessed whether somatic mutations add prognostic value to clinical variables, and whether encoding mutations by their functional impact rather than by binary presence changes which genes are identified as prognostic.Methods: Clinical and somatic mutation data from The Cancer Genome Atlas (TCGA) and the International Cancer Genome Consortium (ICGC) were harmonised into a dataset of 796 patients (discovery: n=597; validation: n=199). Of 43,871 genes, 24 with mutation prevalence ≥5% were analysed. Each gene was encoded either as binary mutation presence (present/absent) or as mutation functional impact (high/low; scored with SIFT, PolyPhen and VEP). Genes were screened with log-rank tests and Benjamini–Hochberg false-discovery-rate correction. Survival was modelled with multivariate Cox proportional-hazards regression and twelve machine-learning (ML) classifiers at 12, 24 and 36 months, using horizon-specific outcomes that excluded patients censored before each horizon. Discrimination on the held-out validation cohort was measured by AUROC (area under the receiver operating characteristic curve) with bootstrap 95% confidence intervals.Results: TP53 was significant after Benjamini–Hochberg correction (q=0.005) and in multivariate analysis. HMCN1 was significant in the multivariate model only under functional-impact encoding (p=0.022; q=0.056) and was not associated with survival under presence encoding. On the held-out cohort, the Cox model achieved an average AUROC of 0.603 and the best ML classifier (AdaBoost) 0.549 (0.638 on cross-validation); both were at chance level at 12 months. Adding mutations to clinical variables did not improve held-out discrimination for either Cox or ML models.Somatic mutation data alone were insufficient for reliable survival prediction in pancreatic cancer and added no discrimination beyond clinical variables. Functional-impact encoding identified HMCN1 as a candidate prognostic factor that requires validation in independent cohorts. Code and data are available on Zenodo.
- 31 May 2025
- Published by Springer Nature
- p. 377-388
Detection Model for Fake News in Bengali Language Using SVM and Random Forest
- 31 May 2025
- Published by Springer Nature
- p. 377-388
Detection Model for Fake News in Bengali Language Using SVM and Random Forest
- 31 January 2025
- Published by Springer Nature
- p. 653-667
Selecting Best DTN Forwarding Protocol on Trace-Based Mobility Models
- 31 January 2025
- Published by Springer Nature
- p. 653-667
Selecting Best DTN Forwarding Protocol on Trace-Based Mobility Models
- 28 May 2023
- Published by Springer Nature
- p. 407-417
Language Identification in Multilingual Text Document Using Machine Learning
- 28 May 2023
- Published by Springer Nature
- p. 407-417
Language Identification in Multilingual Text Document Using Machine Learning
- 2 March 2023
- Published by Springer Nature
- p. 579-591
Remote Health Monitoring System Using Microcontroller—Suitable for Rural and Elderly Patients
- 2 March 2023
- Published by Springer Nature
- p. 579-591
Remote Health Monitoring System Using Microcontroller—Suitable for Rural and Elderly Patients
- 2 March 2023
- Published by Springer Nature
- p. 621-636
Arduino and NodeMCU-Based Smart Soil Moisture Balancer with IoT Integration
- 2 March 2023
- Published by Springer Nature
- p. 621-636
Arduino and NodeMCU-Based Smart Soil Moisture Balancer with IoT Integration