diff --git a/README.md b/README.md
index 09af234dbe370df3ceaaf420bfd0e7f6cda8d3c4..f621b9d245490a2e9245bae173682d46fe3ee1e3 100644
--- a/README.md
+++ b/README.md
@@ -80,12 +80,11 @@ to control the thread pool size used by CAF's scheduler.
---
## Credits
-The implementation of SW-Actors builds on the foundation provided by the [Smith–Waterman algorithm](https://icbi.i-med.ac.at/courses/bioinformatics_ex_2021/7265238.pdf) and the [C++ Actor Framework](https://github.com/actor-framework/actor-framework). We credit the original creators of the C++ Actor Framework which allowed us to implement the actor model into SUMMA-Actors. Links to their research work can be found below.
+The implementation of SW-Actors builds on the foundation provided by the [Smith–Waterman algorithm](https://icbi.i-med.ac.at/courses/bioinformatics_ex_2021/7265238.pdf) and the [C++ Actor Framework](https://github.com/actor-framework/actor-framework). We credit the original creators of the C++ Actor Framework which allowed us to implement the actor model into SW-Actors. Links to their research work can be found below.
* Charousset, D., Schmidt, T. C., Hiesgen, R., Wählisch, M., 2013: Native actors:
a scalable software platform for distributed, heterogeneous environments. _AGERE!_,
[doi:10.1145/2541329.2541336](http://dx.doi.org/10.1145/2541329.2541336).
* Charousset, D., Schmidt, T. C., Hiesgen, R., 2016: Revisiting actor programming in
- C++. _Computer Languages, Systems & Structures_, [doi:10.1016/j.cl.2016.01.002](http://
- dx.doi.org/10.1016/j.cl.2016.01.002)
\ No newline at end of file
+ C++. _Computer Languages, Systems & Structures_, [doi:10.1016/j.cl.2016.01.002](http://dx.doi.org/10.1016/j.cl.2016.01.002)
\ No newline at end of file
diff --git a/benchmarks/README.md b/benchmarks/README.md
index 6581c5d2bae58f2d45257a5b94b58bfbfb032770..9e5cf908372898dc95d09e6ae00ef13c8cd7b5ec 100644
--- a/benchmarks/README.md
+++ b/benchmarks/README.md
@@ -114,5 +114,11 @@ sbatch SWIPE.sh
- Ensure all dependencies (`Parasail`, `SWIPE`, `SeqAN`, etc.) are **installed and loaded** before running.
+## Credits
+The benchmarking setup presented here relies on the work and tools developed by the authors of **Parasail**, **SeqAN**, and **SWIPE**, whose implementations form the basis of our performance comparisons. We gratefully acknowledge their contributions to the bioinformatics community.
+- **Parasail:** Daily, J., 2016. Parasail: SIMD C library for global, semi-global, and local pairwise sequence alignments. *BMC Bioinformatics*, [doi:10.1186/s12859-016-0930-z](https://doi.org/10.1186/s12859-016-0930-z)
+- **SeqAN:** Reinert, K., et al., 2017. The SeqAn C++ template library for efficient sequence analysis: A resource for programmers. *Journal of Biotechnology*, [doi:10.1016/j.jbiotec.2017.07.017](https://doi.org/10.1016/j.jbiotec.2017.07.017)
+- **SWIPE:** Rognes, T., 2011. Faster Smith–Waterman database searches with inter-sequence SIMD parallelisation. *BMC Bioinformatics*, [doi:10.1186/1471-2105-12-221](https://doi.org/10.1186/1471-2105-12-221)
+
---
diff --git a/build/swActor b/build/swActor
index 8c92f96a000eca762c4696224b27b92d3768796d..3cbedd80bf1c712bbe030c4b629af4e57e604f3a 100755
Binary files a/build/swActor and b/build/swActor differ
diff --git a/dataset /BRCA1.fasta b/dataset/BRCA1.fasta
similarity index 100%
rename from dataset /BRCA1.fasta
rename to dataset/BRCA1.fasta
diff --git a/dataset /BRCA2.fasta b/dataset/BRCA2.fasta
similarity index 100%
rename from dataset /BRCA2.fasta
rename to dataset/BRCA2.fasta
diff --git a/dataset /Heterogeneous .fasta b/dataset/Heterogeneous .fasta
similarity index 100%
rename from dataset /Heterogeneous .fasta
rename to dataset/Heterogeneous .fasta
diff --git a/dataset /Titin.fasta b/dataset/Titin.fasta
similarity index 100%
rename from dataset /Titin.fasta
rename to dataset/Titin.fasta
diff --git a/include/pairActor.hpp b/include/pairActor.hpp
index efae9df19139e46b43c4f4e5b3053172050e0085..9dfa7bd7159b670a3315f0a5b3483e829d33ea10 100644
--- a/include/pairActor.hpp
+++ b/include/pairActor.hpp
@@ -22,6 +22,8 @@ namespace caf
std::string alignedSeq2;
actor manager;
int position;
+ int maxLenQuery = 0;
+ int maxLenSubject = 0;
};
behavior pairActor(stateful_actor<pairActorState> *self, int matchScore, int mismatchScore, int gapScore, int deviderRow, int deviderCol);
diff --git a/include/serialActor.hpp b/include/serialActor.hpp
index 9e6060c7c103b789dfaab9c2c82c1e661cfece9f..4e8b42178b29a6835bbc58b271601a7d04aafd77 100644
--- a/include/serialActor.hpp
+++ b/include/serialActor.hpp
@@ -11,6 +11,7 @@ namespace caf
int mismatchScore = 0;
int gapScore = 0;
std::vector<std::vector<int>> scoreMatrix;
+ actor manager;
};
behavior serialActor(stateful_actor<serialActorState> *self, int matchScore, int mismatchScore, int gapScore);
diff --git a/multi.sh b/multi.sh
index 55aa2e16d4d067a06ad48df2ae543d8c00939f1c..54e3da504719441bfc8723846f0c00547cdfdd40 100644
--- a/multi.sh
+++ b/multi.sh
@@ -13,7 +13,7 @@ IP=$(srun --nodes=1 --ntasks=1 hostname -i)
# Run the code
for node in $(scontrol show hostnames); do
echo $node
-export LD_LIBRARY_PATH := /path/to/your/library:$LD_LIBRARY_PATH
+export LD_LIBRARY_PATH= /path/to/your/library:$LD_LIBRARY_PATH
cd $SLURM_SUBMIT_DIR
@@ -23,7 +23,7 @@ export LD_LIBRARY_PATH := /path/to/your/library:$LD_LIBRARY_PATH
if [ $counter -eq 0 ]; then
IP=$(srun --nodes=1 --ntasks=1 hostname -i)
- srun --nodes=1 --ntasks=1 ./swActor -D -s -p 4444 -Q dataset/BRCA1.fasta -A 40 -m 2 -M -1 -g -2 -R 1 -C 1 --caf.scheduler.max-threads=40&
+ srun --nodes=1 --ntasks=1 ./swActor -D -s -p 4444 -Q ../dataset/BRCA1.fasta -A 40 -m 2 -M -1 -g -2 -R 1 -C 1 --caf.scheduler.max-threads=40&
sleep 2
counter=$((counter + 1))
echo "lets do"
diff --git a/single.sh b/single.sh
index fa5a2b122a0103b42035d0504ec8a57ba1a726b1..b1c453b5a95bc0d59d74a9bf261a57e1ff480d11 100644
--- a/single.sh
+++ b/single.sh
@@ -9,7 +9,7 @@
cd $SLURM_SUBMIT_DIR
-export LD_LIBRARY_PATH := /path/to/your/library:$LD_LIBRARY_PATH
+export LD_LIBRARY_PATH= /path/to/your/library:$LD_LIBRARY_PATH
make clean
@@ -20,4 +20,4 @@ cd build
# Run your program
echo "BRCA1 1 node 40 cpu"
-./swActor -Q dataset/BRCA1.fasta -A 40 -m 2 -M -1 -g -2 -R 1 -C 1 --caf.scheduler.max-threads=40
+./swActor -Q ../dataset/BRCA1.fasta -A 40 -m 2 -M -1 -g -2 -R 1 -C 1 --caf.scheduler.max-threads=40
diff --git a/src/actors/clientActor.cpp b/src/actors/clientActor.cpp
index fee4ccbdb879775454f9312643f9ea2eaff83f71..8b9cf8dce6d1ef6843180c387ee10196443bc9e6 100644
--- a/src/actors/clientActor.cpp
+++ b/src/actors/clientActor.cpp
@@ -1,5 +1,4 @@
#include "clientActor.hpp"
-#include "serialActor.hpp"
#include "pairActor.hpp"
namespace caf
@@ -22,22 +21,11 @@ namespace caf
self->state().server = *serverActor;
- if (cfg.deviderRow == 1 && cfg.deviderCol == 1)
+ for (int i = 0; i < cfg.actorNumber; ++i)
{
- for (int i = 0; i < cfg.actorNumber; ++i)
- {
- actor worker = self->spawn(serialActor, cfg.matchScore, cfg.mismatchScore, cfg.gapScore);
- anon_mail(self->state().server).send(worker);
- }
- }
- else
- {
- for (int i = 0; i < cfg.actorNumber; ++i)
- {
- actor worker = self->spawn(pairActor, cfg.matchScore, cfg.mismatchScore, cfg.gapScore,
- cfg.deviderRow, cfg.deviderCol);
- anon_mail(self->state().server).send(worker);
- }
+ actor worker = self->spawn(pairActor, cfg.matchScore, cfg.mismatchScore, cfg.gapScore,
+ cfg.deviderRow, cfg.deviderCol);
+ anon_mail(self->state().server).send(worker);
}
return {};
diff --git a/src/actors/managerActor.cpp b/src/actors/managerActor.cpp
index 71e028fed2316b5146d3fc7354eab84981182f5c..b8a5975b542e6af8dd5d3a75907f1ed7936e646e 100644
--- a/src/actors/managerActor.cpp
+++ b/src/actors/managerActor.cpp
@@ -1,5 +1,4 @@
#include "managerActor.hpp"
-#include "serialActor.hpp"
#include "pairActor.hpp"
#include "makePairs.hpp"
#include "readFasta.hpp"
@@ -48,36 +47,23 @@ namespace caf
self->state().maxLenSubject = self->state().maxLenQuery;
}
- // Start the timer
self->state().start = std::chrono::high_resolution_clock::now();
- if (cfg.deviderRow == 1 && cfg.deviderCol == 1)
- {
- for (int i = 0; i < cfg.actorNumber; ++i)
- {
- actor worker = self->spawn(serialActor, cfg.matchScore, cfg.mismatchScore, cfg.gapScore);
+ self->println("Query sequences: {}", self->state().querySequences.size());
+ self->println("Subject sequences: {}", self->state().subjectSequences.size());
- std::string seq1 = self->state().querySequences[self->state().workList1[i]];
- std::string seq2 = self->state().subjectSequences[self->state().workList2[i]];
-
- anon_mail(self->state().maxLenQuery, self->state().maxLenSubject).send(worker);
- anon_mail(self, i, seq1, seq2).send(worker);
- }
- }
- else
+ for (int i = 0; i < cfg.actorNumber; ++i)
{
- for (int i = 0; i < cfg.actorNumber; ++i)
- {
- actor worker = self->spawn(pairActor, cfg.matchScore, cfg.mismatchScore, cfg.gapScore,
- cfg.deviderRow, cfg.deviderCol);
+ actor worker = self->spawn(pairActor, cfg.matchScore, cfg.mismatchScore, cfg.gapScore,
+ cfg.deviderRow, cfg.deviderCol);
- self->state().start = std::chrono::high_resolution_clock::now();
- std::string seq1 = self->state().querySequences[self->state().workList1[i]];
- std::string seq2 = self->state().subjectSequences[self->state().workList2[i]];
+ std::string seq1 = self->state().querySequences[self->state().workList1[i]];
+ std::string seq2 = self->state().subjectSequences[self->state().workList2[i]];
- anon_mail(self, i, seq1, seq2).send(worker);
- }
+ anon_mail(self->state().maxLenQuery, self->state().maxLenSubject).send(worker);
+ anon_mail(self, i, seq1, seq2).send(worker);
}
+
self->state().position = cfg.actorNumber - 1;
return {
@@ -90,7 +76,7 @@ namespace caf
{
std::string seq1 = self->state().querySequences[self->state().workList1[self->state().position]];
std::string seq2 = self->state().subjectSequences[self->state().workList2[self->state().position]];
- anon_mail(self, self->state().position, seq1, seq2).send(sender);
+ anon_mail(self->state().position, seq1, seq2).send(sender);
}
else
{
diff --git a/src/actors/pairActor.cpp b/src/actors/pairActor.cpp
index ff7b22e2793fcd3e3059c60a18d8632ea6c0abf2..594d74bc668e47b1f9695781cf5dfd9d2b527364 100644
--- a/src/actors/pairActor.cpp
+++ b/src/actors/pairActor.cpp
@@ -1,3 +1,4 @@
+#include "serialActor.hpp"
#include "pairActor.hpp"
#include "blockActor.hpp"
@@ -19,11 +20,25 @@ namespace caf
},
[=](actor manager, int position, std::string seq1, std::string seq2)
{
- actor blockAct = self->spawn(blockActor, actor_cast<actor>(self), 0, seq1, seq2, self->state().matchScore, self->state().mismatchScore, self->state().gapScore, self->state().deviderRow, self->state().deviderCol);
+ if (self->state().deviderRow == 1 && self->state().deviderCol == 1)
+ {
+ actor serialAct = self->spawn(serialActor, self->state().matchScore, self->state().mismatchScore, self->state().gapScore);
+ anon_mail(manager, self->state().maxLenQuery, self->state().maxLenSubject).send(serialAct);
+ anon_mail(position, seq1, seq2).send(serialAct);
+ }
+ else
+ {
+ actor blockAct = self->spawn(blockActor, actor_cast<actor>(self), 0, seq1, seq2, self->state().matchScore, self->state().mismatchScore, self->state().gapScore, self->state().deviderRow, self->state().deviderCol);
+ }
self->state().manager = manager;
self->state().position = position;
self->state().workers.resize(self->state().deviderRow);
},
+ [=](int position, std::string seq1, std::string seq2)
+ {
+ actor blockAct = self->spawn(blockActor, actor_cast<actor>(self), 0, seq1, seq2, self->state().matchScore, self->state().mismatchScore, self->state().gapScore, self->state().deviderRow, self->state().deviderCol);
+ self->state().position = position;
+ },
[=](actor worker, int rowPose, int maxScore, int maxRow, int maxCol)
{
self->state().workers[rowPose] = worker;
@@ -66,7 +81,11 @@ namespace caf
anon_mail(row, col).send(self->state().workers[rowPose]);
}
},
- [=](int maxLenQuery, int maxLenSubject) {},
+ [=](int maxLenQuery, int maxLenSubject)
+ {
+ self->state().maxLenQuery = maxLenQuery;
+ self->state().maxLenSubject = maxLenSubject;
+ },
[=](std::string exit)
{
self->quit();
diff --git a/src/actors/serialActor.cpp b/src/actors/serialActor.cpp
index 944d6aad05a02437bc7afc1eb5b2d5e2d4783f0b..b9072a42beab0073aa05eef50e057b77a9070b1a 100644
--- a/src/actors/serialActor.cpp
+++ b/src/actors/serialActor.cpp
@@ -11,11 +11,12 @@ namespace caf
// receive the messages
return {
- [=](int maxLenQuery, int maxLenSubject)
+ [=](actor manager, int maxLenQuery, int maxLenSubject)
{
self->state().scoreMatrix = std::vector<std::vector<int>>(maxLenQuery + 1, std::vector<int>(maxLenSubject + 1, 0));
+ self->state().manager = manager;
},
- [=](actor manager, int position, std::string seq1, std::string seq2)
+ [=](int position, std::string seq1, std::string seq2)
{
int m = seq1.length();
int n = seq2.length();
@@ -81,11 +82,7 @@ namespace caf
}
self->println("Pair: {}, with score: {}, \nAligned seq1: {}, \nAligned seq2: {}\n", position, maxScore, alignedSeq1, alignedSeq2);
- anon_mail(self, position, maxScore).send(manager);
- },
- [=](actor serverOrManager)
- {
- anon_mail(self).send(serverOrManager);
+ anon_mail(self, position, maxScore).send(self->state().manager);
},
[=](std::string exit)
{