Cognitive biases related to research

Cognitive biases I have commonly encountered in research: Related to hypothesis making/testing,data analyses and decisions for lab experiments

Availability heuristic-overestimate importance of available information
Recency-weigh recent information more heavily
Bandwagon effect-groupthink
Clustering illusion-see patterns in random events
Confirmation bias-confirmation of preconception
Zero risk bias-favoring certainty

For more comprehensive lists

http://mentalfloss.com/article/68705/20-cognitive-biases-affect-your-decisions

https://en.wikipedia.org/wiki/Cognitive_bias#List

Recommended reading: Thinking Fast and Slow-Daniel Kahneman

Assumptions-Are they related to cognitive biases ?

Can we say that an assumption is  a cognitive bias.

Assumption-For what there is no evidence but it is still taken to be true.
Google definition-a thing that is accepted as true or as certain to happen, without proof.

A cognitive bias is a mistake in reasoning, evaluating, remembering, or other cognitive process, often occurring as a result of holding onto one's preferences and beliefs regardless of contrary information. 

So when a person is making a select assumption could that person be unknowingly tapping into their inherent cognitive bias e.g. assumption from memory=availability  heuristic cognitive bias.

Based on this argument "all research" in which each piece of research has some sort of hidden assumptions has inherent cognitive biases of the researcher. 

Fine Graining in the Cancer context

Coarse graining of cellular macrostates can be done at the macroscale. For e.g. if there are cellular macrostates A--->B<=>C---->D, it can be coarse grained to A--->B/C--->D, where A is a group of cancer cells that divide but to a limited extent forming a  pre-cancerous lesion (B) or a benign tumor(C). B and C are interchangeable. When either B or C overcomes the inhibitory signaling of the tumor microenvironment they get converted to a malignant/metastasizing state (D). If  A and D are considered analogous to a  binary switch 0/1 , then the transition between the normal cell and a tumorigenic cell can be considered as :off/on; A/B;0/1 states which can be coarse grained with a single intermediary state B/C.

However each cellular macrostate can consist of  temporally regulated dynamic signal transduction networks. A--->B transition can be considered a series of microstate transitions such as a--->a1--->a2---->a3........b.
Each microstate is a certain configuration of signal transduction networks. For eg a cell with an activating mutation in an oncogene will possess aberrant signaling. A=Normal cell. Oncogenic mutation during replication occurs. Inorder to manifest oncogenic activity the oncogenic protein will initiate and proceed with uncontrolled signaling.
 Microstate a1 is the initiating signaling activity by a single molecule of oncogenic protein. A critical pool of molecules of the oncogenic protein build up slowly say  states a2,a3,a4 , following which sustained signaling by the pool of oncogenic molecules occurs. At various time points the cellular lysates will show different concentrations of the oncogenic proteins.

So a drug candidate can do the following
1) targets oncogenic mutation in state A
2)prevents build up of critical pool of oncogenic protein to 'x' no. of molecules i.e prevents transition from state A
3)prevents signaling by the built pools of oncogenic protein with 'x' no. of molecules i.e.prevents signaling in state B

Depending on the model of tumorigenesis developed by fine graining the best drug candidate can be developed.

Questions -Research Gate

https://www.researchgate.net/profile/Sonali_Sengupta8/questions

Markov chains-Cellular states in Cancer

https://en.wikipedia.org/wiki/Markov_model

Cellular States in Cancer:

Visualization-Oscillating pendulum

Cellular states preceding and following tumorigenesis are  like- an oscillating pendulum that  overshoots mean position once (equivalent to cellular state i.e benign tumor after primary mutational hit-State A). The pendulum tries to  come back to mean balance position (equivalent to restoring homeostasis in cell after primary mutational hit). Due to  loss of elasticity of pendulum string (equivalent to secondary hit due to tumor microenvironment interactions) status quo of extreme position maintained, (which is  equivalent to malignant tumor-State B ).

Idea for a drug- A compound that decreases the probability of  transitioning from state A to state B

Activating a tumor suppressor/inhibiting an oncogene within a tumor by a drug may alter tumor cellular state. Tumor cells may adapt and give survival cues to the tumor microenvironment. So drug has to target signaling by tumor microenvironment to tumor cells. So not one drug but two, targeting tumor cell +tumor microenvironment.

Coarse graining -Disease

Biological Research involves coarse graining-simplifying in order to obtain a model. Is this the reason drugs against diseases like cancer fail in clinical trials ?

Conscious AI

Can Artificial Intelligence become conscious if the information about mental processing which is lost in models during coarse graining is regained ?