Network Analysis for Protein Adaptation

Project Description:

At Johns Hopkins University (JHU), I run a research project with my team member Elysia that studied genetic mutations arising from HIV drug therapies using network analysis and computational tools. For this project, we collaborated with researchers at the JHU Institute for Computational Medicine and presented the final findings to a group of professionals in the field.

Course: Foundations of Computational Biology and Bioinformatics II (Dr. Karchin, Johns Hopkins University – Spring 2018).

Team membersElysia Chou and Marina Melero.

Project Summary:

An Alignment Network-Based Method for Finding Adaptive Mutations due to HIV-1 RT Inhibitors



Research project poster presentation at Johns Hopkins University.

As of 2016, there were 36.7 million people living with HIV/AIDS worldwide with 1.8 million new infections per year [1], and as of yet, there are no effective vaccines for HIV. HIV-1 has been known to be one of the pathogens that result in AIDS for nearly three decades [2]. This pathogen has a high mutant frequency due to its fast-replicating nature. Under the selective pressure of a drug, resistance is quickly acquired by the virus, this constitutes a current prevalent challenge in the development an effective treatment to AIDS.

A better understanding of the genetic mechanisms of drug resistance in HIV-1 would help the development of more effective drugs, and enable doctors to personalize antiretroviral therapy to the needs of each patient. Motivated to gather this understanding, the project examines the use of a Network Analysis tool for Protein Adaptation (NAPA) to identify mutations and mutation combinations that lead to drug resistance to reverse transcriptase inhibitors (NRTIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs) in patients infected with HIV-1.

Project Description:

Longitudinal DNA sequences encoding for the HIV-1 Type B enzyme reverse transcriptase (RT) from 1,713 patients were used to construct a series of alignment-based mutation networks. A filtering and denoising method was developed to clean the networks for analysis. The mutation networks were analyzed using metrics of centrality, modularity, and co-occurrence to identify relevant mutations and interactions. Finally, the results were validated with existing literature and used to demonstrate the efficacy of NAPA in identifying drug-resistance mutations and understanding their mechanism of action.


The open-source software package Network Analysis of Protein Adaptation (NAPA) [3] was used to construct and analyze the undirected alignment-based co-occurrence networks from both the first-time-point and the last-time-point multiple sequence alignments (MSA). Each network file listed mutation pairs and how often these pairs co-occurred.

Screen Shot 2018-10-03 at 6.52.46 PM

Fig.1. NAPA Workflow constructing networks of interactions between mutations based on the MSA on patient’s data.


The following figures (2,3) show the final networks obtained in the analysis. Each node represents a missense mutation with respect to the wild type HIV-1 Type B reverse transcriptase consensus sequence. Node size corresponds to its weighted degree centrality, while the links between the mutations represent the co-occurrences between the respective mutations; link thickness corresponds to the co-occurrence count.

Screen Shot 2018-10-03 at 7.04.59 PM

Fig. 2. Modularity-based network of reverse transcriptase drug adaptation using NAPA and visualized in Cytoscape. NAPA assigned each mutation to one of four communities, as denoted by the different colors.

Screen Shot 2018-10-03 at 7.05.06 PM

Fig. 3. Validation of the network with resistance mutations found in the literature [4].

Relevant mutations were identified, corresponding to the most central nodes and proved to be implicated in drug resistance. In our conclusion, NAPA proved to be a useful and transferable tool to identify mutations linked to drug adaptation, as well as to gain a higher insight on how mutations work interact to develop drug resistance. Our project shows that NAPA can be used for other proteins related to drug resistance across organisms and viruses in order to gain a better understanding of drug resistance emergence which is relevant for both viral pathogenesis and drug development.


  • Full project report available here.
  • Research poster available here.
  • Project presentation slides available here.

Course: Foundations of Computational Biology and Bioinformatics II (Dr. Karchin, Johns Hopkins University – Spring 2018).

Team membersElysia Chou and Marina Melero.


[1] UNAIDS (2016). The Global HIV/AIDS Epidemic [online]. Available at: [Accessed 18 Apr. 2018]

[2] Ibe, S. and Sugiura, W., 2011. Clinical significance of HIV reverse-transcriptase inhibitor-resistance mutations. Future microbiology, 6(3), pp.295-315.

[3] Guthrie, V.B., Masica, D.L., Fraser, A., Federico, J., Fan, Y., Camps, M. and Karchin, R., 2018. Network Analysis of Protein Adaptation: modeling the functional impact of multiple mutations. Mol. Biol. Evol.

[4] Wensing, A.M., Calvez, V., Günthard, H.F., Johnson, V.A., Paredes, R., Pillay, D., Shafer, R.W. and Richman, D.D., 2017. 2017 Update of the Drug Resistance Mutations in HIV-1. Topics in antiviral medicine, 24(4), pp.132-133.






Introduce tus datos o haz clic en un icono para iniciar sesión:

Logo de

Estás comentando usando tu cuenta de Cerrar sesión /  Cambiar )

Google photo

Estás comentando usando tu cuenta de Google. Cerrar sesión /  Cambiar )

Imagen de Twitter

Estás comentando usando tu cuenta de Twitter. Cerrar sesión /  Cambiar )

Foto de Facebook

Estás comentando usando tu cuenta de Facebook. Cerrar sesión /  Cambiar )

Conectando a %s