Sudip Roy

Optimization of Dilution and Mixing of Biochemical Samples Using Digital Microfluidic Biochips

The recent emergence of lab-on-a-chip (LoC) technology has led to a paradigm shift in many healthcare-related application areas, e.g., point-of-care clinical diagnostics, high-throughput sequencing, and proteomics. A promising category of LoCs is digital microfluidic (DMF)-based biochips, in which nanoliter-volume fluid droplets are manipulated on a 2-D electrode array. A key challenge in designing such chips and mapping lab-bench protocols to a LoC is to carry out the dilution process of biochemical samples efficiently. As an optimization and automation technique, we present a dilution/mixing algorithm that significantly reduces the production of waste droplets. This algorithm takes O(n) time to compute at most n sequential mix/split operations required to achieve any given target concentration with an error in concentration factor less than [1/(2n)]. To implement the algorithm, we design an architectural layout of a DMF-based LoC consisting of two O(n)-size rotary mixers and O(n) storage electrodes. Simulation results show that the proposed technique always yields nonnegative savings in the number of waste droplets and also in the total number of input droplets compared to earlier methods.

Waste-aware dilution and mixing of biochemical samples with digital microfluidic biochips

A key challenge in design automation of digital microfluidic biochips is to carry out on-chip dilution/mixing of biochemical samples/reagents for achieving a desired concentration factor (CF). In a bioassay, reducing the waste is crucial because the waste droplet handling is cumbersome and the number of waste reservoirs on-chip needs to be minimized to use limited volume of sample and expensive reagents and hence to reduce the cost of a biochip. The existing dilution algorithms attempt to reduce the number of mix/split steps required in the process but focus little on minimization of sample requirement or waste droplets. In this work, we characterize the underlying combinatorial properties of waste generation and identify the inherent limitations of two earlier mixing algorithms (BS algorithm by Thies et al., Natural Computing 2008; DMRW algorithm by Roy et al., IEEE TCAD 2010) in addressing this issue. Based on these properties, we design an improved dilution/mixing algorithm (IDMA) that optimizes the usage of intermediate droplets generated during the dilution process, which in turn, reduces the demand of sample/reagent and production of waste. The algorithm terminates in O(n) steps for producing a target CF with a precision of 1/2n. Based on simulation results for all CF values ranging from 1/1024 to 1023/1024 using a sample (100% concentration) and a buffer solution (0% concentration), we present an integrated scheme of choosing the best waste-aware dilution algorithm among BS, DMRW, and IDMA for any given value of CF. Finally, an architectural layout of a DMF biochip that supports the proposed scheme is designed.

Layout-Aware Solution Preparation for Biochemical Analysis on a Digital Microfluidic Biochip

A biochemical analysis is based on several laboratory protocols that require repeated mixing of samples with reagents. Sample preparation and analyte identification steps in such bioassays often involve mixing for solution preparation, i.e., various fluids are to be mixed in a certain volumetric ratio in their resulting mixture. We present an efficient approach for automated mixing of three or more fluids on a droplet based digital micro fluidic biochip and design a layout for implementing this algorithm. The proposed method reduces the droplet transportation time from boundary reservoirs to on chip mixers as well as cross-contamination among overlapping droplet routing paths. Simulation of several example solutions reveals encouraging results.

On-Chip Sample Preparation with Multiple Dilutions Using Digital Microfluidics

In many biochemical protocols, solution preparation is a preprocessing step for mixing two or more fluids in a given ratio. Dilution of a biochemical sample/reagent is the special case of mixing or solution preparation where only two different type of fluids, one of which is a buffer solution, are mixed at a certain ratio corresponding to the desired concentration factor. Bioassays implemented on digital micro fluidic biochips may require several different concentration values of the same sample/reagent. In this paper, we present a scheme in which a set of different target droplets (with concentration values ranged between 0% and 100%) can be produced with an acceptable error bound in minimum mix-split steps. The method does not require any intermediate storage since, at each step, the current droplet is mixed only with the sample (with 100% concentration) or with the buffer (with 0% concentration) droplet. The problem of generating multiple target concentrations has been formulated based on a binary de Bruijn graph. The proposed technique outperforms the existing single target based methods in terms of both the number of mix-split steps and the number of waste droplets. This in turn, reduces the execution time, the number of electrode actuations, and sample/reagent requirement. A digital micro fluidic platform can also be easily designed to implement such on-chip sample preparation.

Optimal control strategies for simultaneous saccharification and fermentation of starch

Abstract Design and analysis of optimal control strategies for three types of inhibitory fed-batch bioprocesses have been discussed. These are simple saccharification (SS) of starch to glucose, simple fermentation (SF) of derived glucose to lactic acid (LA) and simultaneous saccharification and fermentation (SSF) of starch to LA. Various optimal feeding strategies have been investigated for the SSF process by manipulating starch addition rates. To avoid the complexity of solving a singular problem, the starch addition rates are expressed in terms of the broth volume, which is used as a control variable. The optimization strategy is thus solved in a nonsingular framework. Experimental studies carried out using the results of the optimization demonstrated the accuracy and utility of the approach. An increase of 20% in lactate productivity was obtained by operating the SSF process in a fed-batch mode. The focus of all the optimization studies has been to improve the performance of the SSF process. Optimal control of starch additions in the fed-batch process gave improved performance of the SSF process.

On-Chip Sample Preparation for Multiple Targets Using Digital Microfluidics

In many biochemical protocols, sample preparation is an extremely important step for mixing multiple reagents in a given ratio. Dilution of a biochemical sample/reagent is the special case of mixing or solution preparation where only two fluids (sample and buffer) are mixed at a certain ratio corresponding to the desired concentration factor. Many bioassays often require multiple concentration values of the same sample/reagent, and implementing them efficiently on a digital microfluidic biochip is a challenge. In this paper, we present an algorithmic solution for the problem of producing a set of different target droplets in a minimum number of mix-split steps, and satisfying a given upper bound in concentration error. Unlike prior methods, this approach does not require any intermediate storage. We represent the underlying search space using a binary de Brujin graph and show that a shortest mix-split sequence can be obtained by solving an asymmetric traveling salesman problem therein. Simulation results over a large data set reveal that the proposed technique outperforms existing methods in terms of the number of mix-split steps, waste droplets, and reactant usage. The method is applicable in general scenarios of either one mixer or more mixers on the chip. A digital microfluidic platform can be easily designed to implement such a technique for rapid on-chip sample preparation.

Efficient mixture preparation on digital microfluidic biochips

Digital microfluidic biochips are recently being developed for on-chip implementation of biochemical laboratory assays. Existing mixing algorithms determine the mixing tree or mixing graph from a given target ratio of several biochemical fluids for on-chip mixture preparation. We present an algorithm to determine a reduced mixing tree by sharing the common subtrees within itself. We observe two transformations that preserve the semantics of the tree: (a) permutation of leaf nodes (input fluids/reagents) within the same level of a mixing tree, and (b) level-shifting of a leaf node to the next lower level by duplicating its appearance. The proposed algorithm utilizes both the intermediate droplets obtained after a split operation when a pair of identical subtrees are identified under permutation of leaf nodes at the same level. Simulation results for a large set of target ratios show that our algorithm reduces the mean values of the total number of mix-split steps, waste droplets and the number of mixer modules required for earliest completion by 16%, 29% and 12% over Min-Mix and by 22%, 34% and 20% over RMA, respectively. Moreover, it reduces the number of checkpoint insertions required for dynamic error recovery against incorrect mix-split steps during mixture preparation.

Layout-Aware Mixture Preparation of Biochemical Fluids on Application-Specific Digital Microfluidic Biochips

The recent proliferation of digital microfluidic (DMF) biochips has enabled rapid on-chip implementation of many biochemical laboratory assays or protocols. Sample preprocessing, which includes dilution and mixing of reagents, plays an important role in the preparation of assays. The automation of sample preparation on a digital microfluidic platform often mandates the execution of a mixing algorithm, which determines a sequence of droplet mix-split steps (usually represented as a mixing graph). However, the overall cost and performance of on-chip mixture preparation not only depends on the mixing graph but also on the resource allocation and scheduling strategy, for instance, the placement of boundary reservoirs or dispensers, mixer modules, storage units, and physical design of droplet-routing pathways. In this article, we first present a new mixing algorithm based on a number-partitioning technique that determines a layout-aware mixing tree corresponding to a given target ratio of a number of fluids. The mixing graph produced by the proposed method can be implemented on a chip with a fewer number of crossovers among droplet-routing paths as well as with a reduced reservoir-to-mixer transportation distance. Second, we propose a routing-aware resource-allocation scheme that can be used to improve the performance of a given mixing algorithm on a chip layout. The design methodology is evaluated on various test cases to demonstrate its effectiveness in mixture preparation with the help of two representative mixing algorithms. Simulation results show that on average, the proposed scheme can reduce the number of crossovers among droplet-routing paths by 89.7% when used in conjunction with the new mixing algorithm, and by 75.4% when an earlier algorithm [Thies et al. 2008] is used.

Theory and analysis of generalized mixing and dilution of biochemical fluids using digital microfluidic biochips

Digital microfluidic (DMF) biochips are recently being advocated for fast on-chip implementation of biochemical laboratory assays or protocols, and several algorithms for diluting and mixing of reagents have been reported. However, all methods for such automatic sample preparation suffer from a drawback that they assume the availability of input fluids in pure form, that is, each with an extreme concentration factor (CF) of 100%. In many real-life scenarios, the stock solutions consist of samples/reagents with multiple CFs. No algorithm is yet known for preparing a target mixture of fluids with a given ratio when its constituents are supplied with random concentrations. An intriguing question is whether or not a given target ratio is feasible to produce from such a general input condition. In this article, we first study the feasibility properties for the generalized mixing problem under the (1:1) mix-split model with an allowable error in the target CFs not exceeding 1 2d, where the integer d is user specified and denotes the desired accuracy level of CF. Next, an algorithm is proposed which produces the desired target ratio of N reagents in ONd mix-split steps, where N ( ≥ 3) denotes the number of constituent fluids in the mixture. The feasibility analysis also leads to the characterization of the total space of input stock solutions from which a given target mixture can be derived, and conversely, the space of all target ratios, which are derivable from a given set of input reagents with arbitrary CFs. Finally, we present a generalized algorithm for diluting a sample S in minimum (1:1) mix-split steps when two or more arbitrary concentrations of S (diluted with the same buffer) are supplied as inputs. These results settle several open questions in droplet-based algorithmic microfluidics and offer efficient solutions for a wider class of on-chip sample preparation problems.

A power-aware wireless sensor network based bridge monitoring system

Recent proliferation of sensor networks in diverse applications has made the low-power wireless sensor network an important design issue. Structural health monitoring is one of the important applications of the wireless sensor network. A robust wireless structural monitoring strategy requires appropriate choice of the wireless network topology as well as some suitable protocols. This paper presents a power-aware sensor network, which can be used for the monitoring of bridges. Event-driven data communication protocols have been used to minimize energy drawn from the battery. All the sensor nodes remain in low-power sleep mode unless triggered either by a base station or by another sensor node in the neighborhood. Two ways of collecting data have been proposed and their performances in terms of the number of motes versus network lifetime have been compared.