K. Satyamoorthy, Ph.D.
Director, Life Sciences Centre, Manipal University, India
Epigenetic Changes due to DNA Methylation in Human Epithelial Tumors
Extensive global hypomethylation in the genome and hypermthylation of selective tumor specific suppressor genes appears to be a hallmark of human cancers. Data suggests that hypermethylation of promoter region in genes is more closely related to subsequent gene expression; contrary to gene-body DNA methylation. The intricate balance between these two may contribute to the progressive process of development, differentiation and carcinogenesis. Epigenetic changes encompass, apart from DNA methylation, chromatin modifications through post-translational changes in histones and control by miRNAs. At the genome level, effects from these are compounded by copy number variations (CNVs) which may ultimately influence protein functions. From clinical perspective, changes in DNA methylation occur very early which are reversible and are influenced by environmental factors. Therefore, these can be potential resource for identifying therapeutic targets as well as biomarkers for early screening of cancer. Our current efforts in profiling genome wide DNA methylation changes in oral, cervical and breast cancers through DNA methylation microarray analysis has revealed number of alterations critical for survival, progression and metastatic behavior of tumors. Bioinformatics and functional analysis revealed several key regulatory molecules controlled by DNA methylation and suggests that DNA methylation changes in several CpG islands appear to co-segregate in the regions of miRNAs as well as in the CNVs. We have validated the signatures for methylation of CpG islands through bisufite sequencing for essential genes in clinical samples and have undertaken transcriptional and functional analysis in tumor cell lines. These results will be presented.
Karmeshu, Ph.D.
Dean & Professor, School of Computer & Systems Sciences & School of Computational & Integrative Sciences, Jawaharlal Nehru University, India.
Interspike Interval Distribution of Neuronal Model with distributed delay: Emergence of unimodal, bimodal and Power law
The study of interspike interval distribution of spiking neurons is a key issue in the field of computational neuroscience. A wide range of spiking patterns display unimodal, bimodal ISI patterns including power law behavior. A challenging problem is to understand the biophysical mechanism which can generate the empirically observed patterns. A neuronal model with distributed delay (NMDD) is proposed and is formulated as an integro-stochastic differential equation which corresponds to a non-markovian process. The widely studied IF and LIF models become special cases of this model. The NMDD brings out some interesting features when excitatory rates are close to inhibitory rates rendering the drift close to zero. It is interesting that NMDD model with gamma type memory kernel can also account for bimodal ISI pattern. The mean delay of the memory kernels plays a significant role in bringing out the transition from unimodal to bimodal ISI distribution. It is interesting to note that when a collection of neurons group together and fire together, the ISI distribution exhibits power law.
Srisairam Achuthan, Ph.D.
Senior Scientific Programmer, Research Informatics Division, Department of Information Sciences, City of Hope, CA, USA
Applying Machine learning for Automated Identification of Patient Cohorts
Srisairam Achuthan, Mike Chang, Ajay Shah, Joyce Niland
Patient cohorts for a clinical study are typically identified based on specific selection criteria. In most cases considerable time and effort are spent in finding the most relevant criteria that could potentially lead to a successful study. For complex diseases, this process can be more difficult and error prone since relevant features may not be easily identifiable. Additionally, the information captured in clinical notes is in non-coded text format. Our goal is to discover patterns within the coded and non-coded fields and thereby reveal complex relationships between clinical characteristics across different patients that would be difficult to accomplish manually. Towards this, we have applied machine learning techniques such as artificial neural networks and decision trees to determine patients sharing similar characteristics from available medical records. For this proof of concept study, we used coded and non-coded (i.e., clinical notes) patient data from a clinical database. Coded clinical information such as diagnoses, labs, medications and demographics recorded within the database were pooled together with non-coded information from clinical notes including, smoking status, life style (active / inactive) status derived from clinical notes. The non-coded textual information was identified and interpreted using a Natural Language Processing (NLP) tool I2E from Linguamatics.