Mr. Ahmed Mehdi


Area: Biostatistics, Computational biology

About me

I am renowned for applying innovative and sophisticated statistical approaches for the translational analyses complex data sets. I have particular expertise in the application of advanced bioinformatic approaches to longitudinal complex data for discovery of biomarkers in type-1 diabetes (T1D) and rheumatoid arthritis. This culminated in a first author paper in JCI Insight where a novel methodology of data correction relative to clinical anchor was suggested. 67 differentially expressed genes were identified that could predict risk of T1D. In another (joint) first author paper in the highly-ranked Sci Transl Med (rank 15/1775), in which novel biomarkers were identified in a clinical trial of antigen-specific therapy using immunomodulatory dendritic cells. In another paper (Clin Transl Immunol), I contributed to develop a list of potential novel drug candidates to delay the T1D.

In my current research, I have developed models of peripheral blood gene signatures and serum biomarkers, resulted in couple of patents (PAT- 2018900128: Methods for determining the risk of developing type-1 diabetes, PCT/AU2014/050415: Kits and methods for the diagnosis, treatment, prevention and monitoring of diabetes). My expertise in applying Bayesian network and linear mixed models to clinically translational outcomes has been further recognized through the recent award of a grant (160K USD) from the JDRF.


Grants and Awards:

  • 2016-2019: Juvenile Diabetes Research Foundation Postdoctoral fellowship, 2016-2019 
  • 2018: Finalist award for Johnson & Johnson Eureka Prize for Innovation in Medical Research (Oscars of Australian science)
  • 2017: Travel award for JDRF Training Grantees Workshop
  • 2016: Juvenile Diabetes Research Foundation Travel Award
  • 2015: Valuing Diversity Grant, department of aboriginal and Torres Strait islander multicultural affairs
  • 2014: Valuing Diversity Grant, department of aboriginal and Torres Strait islander multicultural affairs 
  • 2013: Graduate student international research travel award (GSITA)
  • 2011: Group achievement award IMBcom Bio-business retreat
  • 2009-2013: Research degree scholarship
  • 2008: Research assistantship award 


Supervision (current)

  • Doctor of Philosophy (PhD student) — Principal Advisor

  • Master of  BioTech (MSc student) — Principal Advisor

  • Doctor of Medicine (Postgraduate student) — Principal Advisor


Predicting T1D onset

Type-1 diabetes (T1D) is a chronic autoimmune disease that leads to the destruction and dysfunction of the insulin producing beta cells. T1D affects more than 1.3 million people in Australia at a huge societal cost. The clinical presentation of T1D is preceded by a prodromal period that can last from months to years post birth and is characterised by the production of islet autoantibodies or seroconversion, reflecting loss of immune tolerance to beta cells. The factors that trigger T1D onset are largely unknown, but are believed to be a combination of environmental and genetic cues. Over the last decade, significant advances in T1D research have occurred through studying HLA-high risk individuals at familial risk of T1D into cohorts followed from birth, with concomitant exploration of biomarkers associated with preclinical development of autoantibodies and eventual progression to T1D. Our vision is that a better understanding of T1D progression mechanisms can be established by integrating the huge resource of the heterogeneous data that is available from preclinical studies of T1D development. Such an approach will predict T1D onset and uncover therapeutic strategies that prevent or suppress these pathogenic autoimmune responses in the subjects at highest risk for “disease interception'. Through successful collaborations, we have access to longitudinal microarray data from German BABYDIET, the US DAISY and Finnish DIPP study cohorts from individuals who are at risk of T1D. Some of these individuals progressed to develop T1D. By using this huge resource of data, we aim to find differentially expressed genes in children at risk of T1D. We also aim to link clinical and gene expression data by developing probabilistic models would predict T1D onset.

Required skills: Computer programming, mathematics and biology.

Desired skillls: Interest and experience in Matlab

Reversing Type-1 Diabetes by understanding genomics pathways

Connectivity maps refer to a functional relationship between genes, drugs and diseases. We will utilize reference data of expression profiles from human cancer cell lines treated with FDA approved drugs (Lamb, Crawford et al. 2006, Lamb 2007). This reference data is available at ArrayExpress database (Rocca-Serra, Brazma et al. 2003). By developing a connectivity map of our genes of interest for T1D, we could uncover drugs that could change the expression levels of such genes. Recently, we conducted a longitudinal study to profile gene expressions in peripheral blood (PB) samples from NOD mice at 10 weeks of age, then scored pancreatic insulitis at 14 weeks or determined age of diabetes onset (Pang, Irvine et al. 2015). By using differentially expressed genes, a connectivity map identified 19 drugs which are predicted to induce the ‘protective’ expression profile identified in NOD mice, and thus to delay or prevent diabetes. Of these, 9 overlapped with drugs predicted to induce a ‘non-progressor’ expression profile, based on published human gene expression data. In this proposal we aim to develop connectivity maps for the differentially expressed genes that explain T1D progression in our model.

Through successful collaborations, we have access to longitudinal microarray data from German BABYDIET, the US DAISY and Finnish DIPP study cohorts from individuals who are at risk of T1D. Some of these individuals progressed to develop T1D. By using this huge resource of data, we aim to find differentially expressed genes in children at risk of T1D. We also aim to link clinical and gene expression data by developing probabilistic models would predict T1D onset.

In this project we aim to make drug maps of differentially expressed genes involved in T1D progression. The drug map would reverse the gene expression and consequently the T1D progression.

Required skills: Computer programming

Desired skills: Interest and some experience in R.

Exploring protein abundance dynamics

The abundance of proteins in a cell greatly influence the role they can play in controlling cellu- lar behaviour. To experimentally determine protein abundance, however, remains non-trivial; so predicting it can help us to better understand cellular functions and translational efficiency. Di- verse and large-scale transcriptomics and proteomics data sets that link transcription and trans- lation offer new opportunities to build realistic models of protein abundance. In this chapter, we primarily aim to predict the abundance of nuclear proteins and hypothesize that proteins with similar levels of abundance are co-functional. Moreover, we also aim to predict the abundance of nuclear proteins during cell-cycle phases to investigate whether protein abundance is a better indicator of cell-cycle phases as compared to corresponding mRNA expression levels. 

In this project, I specifically ask two questions: (1) By using expression, sequence and interaction data, can we link transcriptional information with translational data to predict protein abundance better than available methods that assume a linear relationship between such data? (2) Can predicted protein abundance be used to quantify the nuclear proteome in each stage of the cell cycle as accurately as experimentally determined protein abundance? Based on the protein abundance, to what extent can we assign roles to the proteins that are involved in the cell cycle? 

For further details on this project please see the following paper;

Mehdi A, Patrick R, Bailey TL and Bodén, M (2014) “Predicting the dynamics of protein abundance”, Molecular & Cellular Proteomics. 13 (5): 1330-40.

Mehdi A, Ali R, Naqi SA, Ali S, Fatima M, Ishaq M,  Mehdi A, (2019) “Statistical and probabilistic approaches to predict protein abundance”, Ency. Bioinf. Comp. Biol. 3:47-54. 13 (5): 1330-40.

Also have a look into the following software:

PredictPA (Protein abundance predictor)

Required skills: Computer programming, mathematics and biology.

Desired skillls: Interest and experience in Matlab

1) Predicting Type-1 diabetes 2) Developing drugs maps to reverse Type-1 diabetes 3) Inferring nuclear import of proteins 4) Prediciting protein abundance


(ordered by most recent)

1.  Ali R, Naqi SA, Ali S, Fatima M, Ishaq M and Mehdi A (2019) “Statistical and probabilistic approaches to predict protein abundance" Ency. Bioinf. Comp. Biol.3:847-854.

2.  Zahir SF, Griffin A, Veerman JL, Magliano D, Shaw JE, Le Cao KA and Mehdi A (2019) “Exploring the association between BMI and mortality in Australian women and men with and without diabetes: the AusDiab study” (accepted Diabetologia, IF=6.2, January 2019)

3.  Rehaume L, Matigian N, Mehdi Aet al., (2019) “IL-23 favors outgrowth of spondyloarthritis-associated pathobionts and suppresses host support for homestatic microbiota" (accepted Annals Rheumatic Diseases IF=12.35, January 2019).

4.  Zahir SF, Diomedi BZ, Mehdi A and Downes M (2019) “Examining the obesity-mortality association in individuals with hypertension: A systemic review and meta-analysis” (under review)

5.  Garcia-Gonzalez, P, Maggi J, Schinnerling K, Soto L, Neira O, Sepulveda-Gutierrez A, Mehdi A et al., (2019) “Regulation of tolerogenic features on dexamethasone-modulated MPLA-activated dendritic cells by MYC” (under review)

6.  Kerry B, Mehdi Aet al., (2019) “A clinical model incorporating factors related to insulin sensitivity predicts stimulated C-peptide in children with type 1 diabetes" (accepted in Pediatric Diabetes, December 2018),

7.  Mehdi A, Hamilton-Williams E, Cristino A, Ziegler A, Bonifacio E, Le Cao KA and Thomas R(2018) “A peripheral blood gene expression signature to predict autoantibody development in infants at risk of type 1 diabetes" JCI Insight. 3(5): e98212.

8. O’Sullivan BJ, Yekollu S, Ruscher R, Maradana MR, Mehdi A, Chidgey AP, Thomas R(2018) “Autoimmune-mediated thymic atrophy is accelerated in RelB-deficient mice" Front. Immunol.9(1092): 1-11

9.  Bhuyan ZA, Rahman MA, Maradana M, Mehdi A,et al., (2018) “Intestinal Inflammatory Regulation in Murine and Human Spondyloarthropathy requires high-affinity T cell receptor-zeta chain-associated protein (ZAP) 70-mediated runx related transcriptional activity. Arthritis Rheumatol.70 (1): 2026-2027.

10.  Ison M, Duggan E,Mehdi AThomas R and Benham H (2018) “Treatment delays for patients with new onset rheumatoid arthritis presenting to an Australian early arthritis clinic" (Accepted: 21-5-2018, Intern. Med. J

11.  Ali R, Mehdi A, Henning S, Marsh B, Landsberg M and Hankamer B (2017) “RAZA: A Rapid 3D z-crossing Algorithm to segment electron tomograms and extract organelles and macromolecules" J. Struc. Biol. 200(2): 73-86.

12. Garcia-Gonzalez, P, Schinnerling K, Sepulveda-Gutierrez A, Maggi J, Ubilla-Olguin G, Mehdi A et al., (2017) “Dexamethasone and monophosphoryl lipid A induce a distinctive profile on monocyte-derived dendritic cells through transcriptional modulation of genes associated with essential processes of the immune response”, Front. Immunol. 8 (1350): 1019.

13. Garcia-Gonzalez, P, Schinnerling K, Sepulveda-Gutierrez A, Maggi J, Hoyos L, Morales R, Mehdi A et al., (2016) “Treatment with dexamethasone and monophosphoryl lipid A removes disease-associated transcriptional signatures in monocyte-derived dendritic cells from rheumatoid arthritis patients and confers tolerogenic features”, Front. Immunol.  7 (43): 1-11.

14. Law SC, Nel HJ, Scally SW, Mehdi A, et al., (2016) “Oligoclonally expanded CD4+T cells recognising citrulinated vimentin in peripheral blood of rheumatoid arthritis patients”, Eur. J. Immunol.46 (S1): 933-933. doi:10.1002/eji.201670200

15. *Benham H, *Nel H, *Law S, *Mehdi A, Street S, Namnoruth N, Pahau H, Lee B, Ng J, Brunck M, Hyde C, Trouw L, Dudek N, Purcell A, O'Sullivan B, Connolly J, Paul S, Cao KA, Thomas R (2015) “Citrullinated peptide dendritic cell immunotherapy in HLA risk genotype-positive rheumatoid arthritis patients”, Sci. Transl. Med.  7 (290): 1-11. (* = equal contribution as first authors,IF=16.8

16. Pang D, Irvine K, Mehdi A, Fynch S, Thomas H, Harris M, Hamilton-Williams, E and Thomas R. (2015) “Peripheral blood gene expression profiling of pre-diabetic NOD mice uncovers novel clinical biomarkers with clinical and therapeutic potential for type-1 diabetes”, (2015). Clinical and Translational Immunology. 4 (8): 1-9.

17.  Mehdi A, Patrick R, Bailey TL and Bodén, M (2014) “Predicting the dynamics of protein abundance”, Molecular & Cellular Proteomics. 13 (5): 1330-40.

18. Róna G, Borsos M, Ellis J, Mehdi A, Christie M, Környei Z, Neubrandt M, Tóth J, Bozóky Z, Buday L, Madarász E, Bodén M, Kobe B and Vértessy B (2014) "Dynamics of re-constitution of the human nuclear proteome after cell division is regulated by NLS-adjacent phosphorylation", Cell Cycle. 1-43.

19.  Mehdi A, (2013) “Computatioanl models of nucleo-cytoplasmic trafficking by integrating heterogeneous data. PhD thesis, The University of Queensland, 1-206.

20.  Mehdi A, Sehgal S, Kobe B, Bailey TL and Bodén M (2013) “DLocalMotif: A discriminative approach for discovering local motifs in protein sequences”, Bioinformatics, 29 (1): 39-46.

21. Mehdi A, Sehgal S, Kobe B, Bailey TL, Bodén, M (2011) “A probabilistic model of nuclear import of proteins”, Bioinformatics, 27:1239-1246

22. Marfori M, Mynott A, Ellis J, Mehdi A, Saunders NF, Curmi P, Forwood J, Bodén M and Kobe B (2011) “Molecular Basis for specificity of nuclear import and prediction of nuclear localization”, Biochim Biophys Acta, 813(9):1562-77.

23. Mehdi A,Zayegh A, Begg R, Ali R (2010) “GK based Fuzzy clustering for the diagnosis of cardiac arrhythmia”, International Journal of Computational Intelligence and Application, 9(3): 105-123.

24. Raza, T, Sheikh, N, Fahiem A, Mehdi A(2011) “Performance of different approaches for predicting the subcellular locations of proteins: A review”, IEEE Proc. : 90-95.

25. Mehdi A, Sehgal M, Zayegh A, Begg R, Manan A (2009) “K-Means Clustering on 3rd order polynomial based normalization of Acute Myeloid Leukemia (AML) and Acute Lymphocyte Leukemia (ALL), IEEE Proc. : 1-5.

26. Mehdi A, Zayegh A, Begg R (2010) “Analysis of Spastic Diplegia Form of Cerebral Palsy Gait using GK based Extended Fuzzy Clustering Approach”, Proc.  Mod.  Sim.

27. Paracha Z, Mehdi A, Kalam, A. (2009) “Computational analysis of sag and swell in electrical power distribution network” IEEE Proc. : 1-5.

28. Paracha Z,, Kalam, A, Mehdi A, Amanullah MTO (2009) “Estimation of power factor by the analysis of power quality data for voltage unbalance” IEEE Proc. :1-4.


Research fields

Bioinformatics, Biostatistics, Computational Immunology


ResearcherIDORCIDGoogle Scholar