Neural stimulation techniques for eliciting calcium influx can elucidate the physiological roles of specific neural populations. To overcome some of the limitations of existing techniques such as poor specificity and noxious effects of heat, I developed a technology for non-invasive control of neural activities using magnetic forces and magnetic nanoparticles (MNPs) which offer deep tissue penetration and controllable dosage. Extensive investigations with different neurotoxins and experimental conditions support that the mechanism of magnetic stimulation that involves membrane-bound MNPs transducing magnetic forces into mechanical stretching of the cell membrane to enhance the opening probability of mechano-sensitive N-type calcium ion channels to induce calcium influx.Making use of the ability of neural networks to actively regulate their ratio of excitatory to inhibitory ion channel/receptor, I also performed chronic magnetic stimulation on fragile X iii syndrome (FXS) model neural networks. We found that chronic magnetic stimulation reduced the density of N-type calcium ion channels whose expression is increased in FXS. This technique demonstrates the potential of using bio-magnetic/mechanical forces to modulate expressions of mechano-sensitive ion channels where they are over-expressed in diseases such as abnormal nociception.Nonetheless, there are still a few areas where the technique can be improved. Firstly, it is the use of MNPs with more uniform properties to have greater control on magnetic stimulations.Secondly, the technique needs to be useful for in vivo studies. Therefore, I started researching on magnetotactic bacteria (MTB) which produce biological MNPs with superior properties such as uniform sizes and highly homogenous magnetic properties with the goal of harvesting MNPs from them.MTB, however, grow extremely slowly and the number of MNPs produced/bacterium is low. One way to overcome this problem is to evolve MTB over-producers of MNPs but this strategy is constrained by the absence of a selection platform that is quantitative and offer high throughput. To overcome this problem, I combined random chemical mutagenesis and selection using a magnetic ratcheting platform to generate and isolate MTB over-producers that produce twice as many MNPs/bacterium after 5 rounds of mutation/selection. I next designed a magnetic microfluidic device and demonstrate as a proof of concept, that it can be coupled to a bioreactor for high throughput microfluidic selection of MTB over-producers.iv