Hardware Reference
In-Depth Information
Tabl e 5. 1
Operations and
their corresponding regions to
be performed
Operations
Corresponding region to be performed
Subtree
M
8
Region
.R
4
;R
7
/
Subtree
M
9
Region
.R
1
;R
4
/
Subtree
M
13
Region
.R
2
;R
6
/
Subtree
M
14
Region
.R
3
;R
5
/
M
12
Region
.R
4
;R
7
/
M
15
Region
.R
4
;R
7
/
M
16
Region
.R
3
;R
6
/
M
17
Region
.R
4
;R
7
;R
2
;R
6
/
M
18
Region
.R
2
;R
7
/
5.4.3
Bioassay-Specific Reservoir Allocation
Based on above discussion, the results for bioassay-specific reservoir allocation,
droplet routing, and the scheduling of fluid-handling operations can be derived as
follows.
1.
Clustering of sequencing graph
: As discussed in Sect.
5.4.2
, if the operations
in a subgraph
SubG
only involve droplets from two reservoirs, then all the
operations in
SubG
can be performed in the region between these two reservoirs.
Therefore, in this stage, we need to find the maximal subgraph in the sequencing
graph, such that the operations in the subgraph involve only two reagents/samples
RS
i
and
RS
j
. This subgraph is defined as “basic mixing clusters” (written as
cluster
.RS
i
;RS
j
/) in the sequencing graph.
Starting from the nodes that represent the final mixing operations of the bioas-
say (i.e., the nodes of zero out-degrees), we check the number of “components”
of its output droplets. For example, the output droplet of operation M
1
8 is a
mixture of four reagents R
1
, R
2
, R
3
,andR
4
. Then we need to trace back to the
predecessor operations of M
1
8, and then check the components of their output
droplets. This trace-back procedure will repeat until finding the operation N
that the corresponding output droplets have no more than two components. Then
the subgraph consisting of operation N
and its predecessor operations is a basic
mixing cluster in the sequencing graph.
2.
Reservoir allocation:
After finding out all the basic mixing clusters of a
sequencing graph, the bioassay-specific reservoir allocation can be determined.
For each mixing cluster
cluster
.RS
i
;RS
j
/, the reservoirs corresponding to
RS
i
and
RS
j
will be placed as the “neighbors” on the layout. The concept of
neighbors on the electrode ring is defined as below:
Definition 1.
Assume that there is a set of reservoirs S
R
Df
R
n
1
;R
n
2
;R
n
3
;:::;
R
n
k
g
, and all these reservoirs are placed on an electrode ring. A pair of reservoirs
R
n
i
and R
n
j
(n
1
n
i
<n
j
n
k
) are “neighbors” in S
R
if there is an electrode
path P
.R
n
i
;R
n
j
/
goes through the output ports of R
i
and R
j
, and there is no output
port of other reservoirs contained by the set S
R
on this electrode path.
