BioKG Portal: A Biomedical Knowledge Graph for Gene-phenotypical Analysis Service in the Domain of Cancer

Abstract

Technology is evolving very fast and many complicated activities are becoming simple to do. Similarly, many technologies are available in market which can used to improve patient care. However, it does not mean that what is done is enough for us. Even in this era we face lot of problems in health or medical industry although health industry growth day by day but still we need lot of improvement in the available technologies and lot of diffract techniques to make our health better. The focus of out project on Cancer which is a very deadly disease. Tassociation are many application and web portals available for helping Doctors, Bioinformatician and other researchers to make evaluate and analyse the different data to get better knowledeg about the core issue of disease Like cBioportal.Our Portal is to improve patient care and enhance the collaboration among patients, care givers and healthcare practitioners (doctors). The proposed system “BioKG Portal” facilitate Clinicians and Bioinformaticians directly and Patients indirectly.By this proposed System, the Clinician (Doctors) can first load two types of data (Clinical Data and Genomic Data) which will help them to analysis the data easily. Moreover, the user can visualise these types of data with the help of different Graphs and in future we will extend it to the descriptive summarization with reasoning ,association Data visualisation can done on each data set like clinical data and Genomic data separately. The graphs included Line plot, Bar plot, Box-plot, Dist plot, Violin plot and Scatter plot. Graphic Visualisation will help the user to easily visualise the data, because the data is too much large, and it is difficult for user to analysis the data manually.Moreover we know that Many doctors are doing treatment of patient on the basis of tassociation experience and knowledge which they have so association is chance of mistake is high and they can’t fine correlation/association between clinical data and genatic data because the dataset are very large and tassociation are thousands of genes in human body so every doctor have not knowledge of every gene but on the other hand side when the doctor analyse the genatic data and clinical data by using software and find the association between clinical data and genomic data then the chance of mistake is low. For this first, we must prepare two types of data files, Phenotypic data file and Expression data file. The system will generate graphic results and the user can visualise the results easily. We have achieved the data association using GSEA method. The data association results will help the User and researcher to extract new knowledge to improve their research work. The system will also integrate the Grakn to generate knowledge graph. The knowledge graph will help to produce new knowledge from clinical data and genomic data. To implement knowledge graph, we must first make schema, then we feed data to schema. After that we have define some rules under some medical techniques and association techniques of genes and clanical techniques. and under certain define rules,and when we query to system then the system will extract knowledge from data under defined rules.