Database Integration Model

Bioinformatics data is continuously increasing day by day. There are different databases that keep the information about genome sequence, variants and genes. But all of them have different format and not user-friendly for end user. User could not easily filter the relevant information and also the information on all databases is heterogenous. Our project is to design a model that integrates the data from all databases and easily filter the data at time retravel. All the data will be accessed and stored in our schema by writing the parsers. At the end, we will provide user friendly interface and filter data with respect to user’s query and solve the problem of heterogeneity.

Cancer Classification

Cancer classification is the idea of the predicting the origin where cancer was born. Our project will predict the type of cancer like brain or lung cancer by human genome data. It will help in fast, timely and accurate discovery of cancer type. Also, the appropriate drugs can be predicted after valid identification of cancer type. Cancer driver genes will be predicted and a classifier will be trained on driver genes. At the end, we will be able to predict the cancer type by knowing the information about few genes.

Protein-protein interaction and visualization

Protein–protein interactions are the physical contacts of high specificity established between two or more protein molecules as a result of biochemical events. We are working on protein-protein interaction, protein complex identification, identification of cancerous and other disease-causing proteins and making it automatic using machine learning techniques. We are also working on the 3d-visualization of protein structure and the interaction that will help to understand the structure and functionalities of proteins much efficiently.

Drug Discovery

Despite the significant achievements in chemotherapy, cancer remains one of the leading causes of death. Target therapy revolutionized this field, but efficiencies of target drugs show dramatic variation among individual patients. Personalization of target therapies remains, therefore, a challenge in oncology. We are working on drug discovery process that will help to recommend the drug according to mutation rate in patient. Mutation rate helps to identify the stage of disease. Besides that, we are also working on drug-gene association prediction that will help to know that how much a drug will affect a particular gene.

Genome contamination identification and removal

We are dealing with contamination in DNA sequences. Contamination is change in the sequence, it can be insertion, deletion, or updation of bases due to many causes. The major causes of contamination are environment, reagents, handlers, or machines at any point during the collection of the sample, the extraction of nucleic acid, or the preparation of libraries. We are identifying and removing different type of contamination in human genome.

Gene Sequencing

Sequencing DNA means determining the order of the four chemical building blocks – called “bases” – that make up the DNA molecule. The sequence tells scientists the kind of genetic information that is carried in a particular DNA segment. For example, scientists can use sequence information to determine which stretches of DNA contain genes and which stretches carry regulatory instructions, turning genes on or off. In addition, sequence data can highlight changes in a gene that may cause diseases. We are working on genome sequencing to identify variation and contamination in genes. Our aim is to discover drugs and identify disease on the basis of genome sequencing.

Cancer Subtype Identification

Cancer is a genetic disease that have different stages. We consider the stage one cancer as initial stage cancer, stage 2 as middle stage and stage 3 or 4 as last stage cancer. It was very crucial for the drug diagnosis to find the patient cancer stage. Cancer subtype identification predicts the cancer stage of the patient that would be initial, middle or last stage. It supports to timely discover the patient stage and therapy of disease.

Visualize Different species in Phylogenetics Tree

Phylogenetic is the process of defining the evolutionary relationship among different species like human, monkey and bee etc. Phylogenetics process creates the phylogenetic trees that show the relationship between species as well as the different samples of single species. All nodes that are on same branch would be considered as closely related to each other. Our tool calculates the distance between different species by using DNA sequence. Its purpose is to show the evolutionary distance of different entities. Different visualization method are also provided to show similarities and dissimilarities between species.

Robot navigation with brain activity of facial expressions and eye movements

This application is based on identifying facial expressions and eye movements through EEG signals. The facial expressions such as smile, sad, teeth clench, smirk, and wink are classified.  Classification based control signals were then transmitted to robot for navigation. The robot is based on shared control which is safe and robust. The analysis of robot navigation for patients showed promising results.

1.    Brain controlled prosthetic arm

This application employs brain signals related to various cognitive states of a user. Cognitive states such as user’s focus level, stress level, engagement level, excitement level and interest level are classified. Based on the level of particular brain activity, prosthetic hand is controlled.

2.    Brain controlled Quadcopter

This application uses brain signals related to imagined movements (motor imagery) which are classified and are used to control the quadcopter.

3.    Machine Learning Tool Kit for BCI

We have developed a toolkit for Machine Learning (ML), which is specifically designed for the purpose of processing and classification of raw EEG data. Moreover, this tool kit can be used for examining various machine learning algorithms on any EEG datasets.

Machine learning toolkit

Brain informatics: Modeling and classification of brain data

Brain Computer Interfacing (BCI) is changing the way humans communicate with machines. The initial interest in BCI emerged in biomedical domain to restore the motor functionalities of locked-in patients – the people who suffer from serious physical disabilities. The total population of persons with disabilities (PWDs) is about 5 million in Pakistan.

Our project is to develop Brain Computer Interface (BCI) for wheelchair navigation through Electroencephalography (EEG) for severely paralyzed people. Our aim is to investigate the modeling and classification of brain signals through EEG. These signals will be modeled, classified and commands based on the classification will be sent to the wheelchair. Such system will provide provision to disabled people in daily life and in their professional life. With this precious gift of mobility, disabled people can start a new life in new ways and contribute better with family and friends.

Physically impaired and disabled people are an integral part of human society. Devices providing assistance to such individuals can help them contribute to the society in a more productive way. Assistive robot, prosthetic hand, and quadcopter controlled by BCI are some of the applications of assistive devices. For this purpose, EEG headset is used to record electroencephalogram (EEG) signals for classification and processing.  Classified control signals are then transmitted to device for performing intended action. We have tested various BCI paradigms on robotic applications