In the awake of Google’s sister outfit, alphabet decided to venture into treating psychological problems such as schizophrenia Autism, bipolar disorders, depression etc. to name a few and it comes at the juncture, they managed to make a coup of recruiting Dr. Thomas Insel as the head, who will spearhead the initiative. I thought it will be none other better time to post my views though I am late for a month to air my views on what I think the research problems we face when we tried to subdue the demons in the brain using very simple molecular tools. In spite of being spent gargantuan amount of money every year to cure depression and mental health diseases, we haven’t yet managed to find either a partial cure or zeroed down the causality of the disease. I would argue it is also partially because many of the molecular researchers consider brain as a cell and their reductionist approach of pinning down a single gene responsible for a disease or a syndrome acts as a stumbling block of further progression. Does that happening because we still don’t have the tools to address more than a two-body problem so that we can reduce the variability in our experiments to make conclusive statements of at least what we observe in the rigid controlled experimental paradigm? or we believe this reductionist observation holds true when it need extrapolating into a holistic view of how brain functions? How do we equate those observations to a bigger view of how brain functions given each neurons can make thousands of connections culminating in trillions of synapses across different regions of the brain where different areas cross-talk to each other. Though there are shortcomings of doing molecular neuroscience to understand the role of genes in the psychiatric diseases, with no reservations, I strongly believe that molecular research is important to dissect the molecular mechanisms of how a neuron operates or how this has been affected in a challenged state. Besides, molecular neuroscience plays a pivotal role in dissecting the subtypes of neurons using molecular markers that can be a valuable asset for genetic tricks to manipulate sub-population of neurons. By using molecular tools will accelerate to help us in manipulating population of neurons and combining with tools of electrical and optical methods, we are in a unique state to manipulate neural activity that can be translated into experience dependent behavioural responses. However, using molecular approach as only as means to understand how brain process, consolidate and retrieve information and how this has been comprised in a diseased state is a far-fetched out cry. We need taking into consideration that neurodegenerative diseases are different from psychological diseases such as autism, depression schizophrenia, or bipolar disorders for instance. This is also due to the fact that in the later cases, not only the nature such as genes plays a role but also nurture does a prominent role in disease progression.
In the midst of the plethora of wealth of information from a gamut of fields ranges from genomics to transcriptomics to single cell analysis to genome wide association studies to comparative genomics studies from cells to model organisms to patients deposited in multiple public databases increasing far beyond the capacities of analysing them in the laboratories. The complex tasks for developing rigorous statistical and computational methods are in an urgent need of hour to extract meaningful information. The foray of Google, the data giants stepping into working on Psychiatric problem will be a great catalyst to accelerate the process of data mining to hopefully find more meaningful information from big data sets eventually might offers clue to find proper therapeutic interventions. I believe this will be the case. To warrant my statement, I am enumerating a few arguments of shortcomings of solely using molecular data as a proxy for learning about neurocognitive disorders but very strongly advocate that besides molecular studies its foremost important to combine comparative genomics with genome wide association studies to identify multiple variants including the rare ones, elucidating the structural variants such as insertions and deletions to have a comprehensive understanding of the genetics of the cognitive disorders so that we might be in a vantage position to rationale therapeutic interventions.
- One of the foremost thing is that neuropsychiatric diseases won’t be considered as a molecular disease such as one gene one disease hypothesis.
There are many kinds of mutations (low, medium, high risk), which render the person to pre-dispose to diseases such as Autism and Schizophrenia, though the existence of low and medium risk mutants doesn’t guarantee that the person will suffer from above diseases. Most of the work till dated, which tried to understand the molecular genetics behind the neurodevelopmental diseases comes from single nucleotide variants (SNV’s) and from copy-number variants (CNV’s). This work has laid foundation in understanding the common and rare genetic variants which causes this neurocognitive disorders. However we were not able to identify a single (SNP’s) single nucleotide polymorphism from genome –wide association studies, which already points out that multiple loci’s get affected in patients either primarily or a consequence of primary and secondary changes owing to the neural activity, history of activity etc. In this context, I believe systematically combining the medley of large-scale comparative genome sequencing to delve on the structural changes (insertions, deletions) or variations (SNP’s) in patients and in healthy subjects with diligent experiment approaches will eventually acts as a scaffold for discovering the genetic architecture of the diseases
- It is obvious and makes sense many of the genes which mutates causes neurocognitive disorders are key molecules involved in synaptic transmission, scaffolding proteins in keeping the synapse architecture, activity dependent genes in maintaining neuronal homeostasis to modulate network activity. However, increasing in the number or decrease in the synapse density as a proxy for rationalizing your favourite gene as a causality for the spectrum of disorder is misleading and not true. One need taking into consideration and often neglected is that the behavioural phenotype is observing is due to the multitude of effects. Gross changes in synapse number, density might be caused by a series of effects produced by original mutations, a series of secondary mutations augmented by secondary changes in neuronal activity, morphological changes or some other compensatory mechanisms.
- The famous saying “ All roads leads to Rome” in other words, you can say, you can take multiple roads to reach Rome. Along the same line, you can have neurocognitive disorders through multiple pathways. Our brains have evolved or adapted in such a way that there is a tremendous amount of redundant pathways, tinkering them will or will not lead to a similar set of behavioural phenotype. Pushing hard your favourite gene, which pops up one among the thousands in a genome wide associations studies of autism or schizophrenia risk factors shows that there is a problem of conceptual frame work in having a clear vision of how to address the intangible problem of how your brain states changes upon cognitive disorders. It is worthwhile to note it becomes a common practise to extrapolate the single gene studies to complex neuropsychiatric disorders. Moreover we have to bear in mind that neurocognitive disorders cannot be considered as molecular disorders. We need knowing not only the molecule A or B but knowing the system components of how the circuits works, what kind of operations are made in circuits or in the networks? What kind of changes happened in network output caused by the local changes (synapse numbers, strength, density) in the microcircuits and how this network has been compromised in cognitive disorder? It is important to study the weight or strength of the synapses and the mechanism affecting the changes: in nutshell we need upgrading ourselves to asking questions from molecular level to system level with a rigorous methodological approaches.
- Caution has to be taken when we try to use mouse models to understand the causality of dysfunction of genes in neurophysciatric disorders. The diversity of this disorders points out that there are different mutations, which can be linked for the diversity, and finding a far-stretched nexus between a gene and the disease thereby might not shed solutions for therapeutic interventions. Given the number of behavioural read-outs for cognitive disorders for autism, schizophrenia, bipolar disorder is limited in mouse models, translating this data in comparison to that of the human data may lead to many artefacts. Discovering new “ disease like” phenotypes solely based on behavioural paradigms but without knowing the epidemiology and the heritability of the diseases will add up more difficulties for interpretation.
- Not only nature, the genetic element but also nurture has a major role in pathophysiology of the disease. We cannot ignore the environmental factors such as deprivations, abuses, and experiences in childhood and adulthood’s etc., shapes our cognitive behaviours. Moreover the genetic backgrounds possess remarkable ability to buffer the deleterious mutants, therefore the susceptibility to cognitive disorders varies between individuals, which will be hard to replicate in mouse models. Besides the synaptic homeostasis can keep the imbalance in network activity in check. The proponent view of setting the imbalance in excitation and inhibition to normal levels doesn’t always hold true as the diseased state is manifested both by nature augmented extensively by nurture.
- I believe it will be interesting to see how the connectomics of the mouse models for this spectrum of disorders has been changed to that of the control animals. With the advent of new tools and techniques we are able to start mapping the connectome for correlating structures to function. The optical clearing methods also enable to see coarse details of how the connectome of the mutants changes to the normal one. Though this will not be able to help us to extract information on the weight of individual synapses, it still gives anatomical differences in wild type and in mutant if any in a coarse level. Knowing the convergence and divergence of connectivity is a crucial factor that determines optimal neural code for information transfer between two networks. In neural systems there is an interaction between network structure and network activity as defined by activity plasticity rules. Through the approach of connectomics, we will be able to glean more information on the information transfer between networks in control and in mutants.
- Patients suffering from neuropsychiatric diseases have to be taken care properly. There are many non-profit foundations to take care, help children and patients suffering from cognitive disorders and also spread awareness among public how to help the patients and integrate them in our normal world. Technology can play its role in helping this patients, one of them which caught my attention was the crowd sourcing tools for anxiety and depression (http://news.mit.edu/2015/crowdsourced-depression-tool-0330 ) where they build on-line support communities and use therapeutic techniques such as cognitive reappraisal to improve the mood of the subjects. I hope, Google can build on-line tools which can help improve the mental states of the patients who been suffering from neuropsychiatric disorders.
- If Google venture into large scale genome sequencing and data analysis, it will hasten the sample size by many folds and help to develop powerful algorithms to analyse and extract meaningful data
- Whole genome, large-scale exome analysis will tells us about the number and frequency of the genetic variants.
- Genome wide association data on families affected with ASD, Schizophrenia, bipolar disorder etc. as well as population-based studies will shed lights on the co-occurrence of SNP’s and rare variants associated with the particular disease
- Comparative genome sequencing and data mining across population will tell us the diversity of phenotypes observed in populations and across populations. We might also glean information on why certain genetic backgrounds are more susceptible to high risk to the disease compared to the other backgrounds, in other words what genetic variants in the genome render strong genetic buffering ability to negate the deleterious efforts of the mutation. Collating information from the family trees can reveal those genetic buffering backgrounds as well as figuring out the highly penetrant mutation.
- Comparative genome sequencing will teach us how structural changes such as CNV’s, SNP’s are involved in cognitive disorders and how these genetic variants are selected in a population, elucidating hot spots in the genome (till date there is no preference for a specific loci). Applying quantitative trait loci mapping studies on patients, their families, in population studies will give us in a vantage position of measuring epigenetic changes to the transcriptome. Combining rigorous statistical and computational methods will give us an opportunity to use magnifying lens to draft hypothesis from genome, epigenome and transcriptome data sets from thousands of patients and their families.
- Mapping how functional regulatory variants influence the expression of synaptic proteins in neurocognitive disorders and their role in the etiology of the disease will be an important approach to be taken. It will be interesting to see how the deleterious variants are selected to adapt in population.
We are currently in a situation of reaping benefits of multiple studies coming from insertion/deletion, SNP’s, genome wide association from individual labs and also from consortiums, for instance recently the 1000 genome data has been published, which offers a plethora of wealth of data of information on genetic variations within human population dispersed across 5 continents. Google’s foray into understanding on mental diseases will complement and hasten already much appreciated work from many labs around the world to eventually streamline a proper blue prints to rationale therapeutic intervention for many neurocognitive diseases. I am optimistic!