Recent changes to Home

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Fri, 17 Aug 2012 20:18:28 -0000

--- v20
+++ v21
@@ -2,8 +2,8 @@
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
-Last update: July 23, 2012
-Improved speed for the assembly and improved performance of the scaffolder of the beta version.
+Last update: Aug 17, 2012
+Source code for the beta version is released. K-mer counting is enabled. Several bugs fixed. 
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1
@@ -18,7 +18,12 @@
 
 Updates: 
 
-Last update: July 8, 2012
+
+Last update: Aug 17, 2012
+Source code for the beta version is released. K-mer counting with the sparse k-mers is enabled. Several bugs fixed. 
+
+
+July 8, 2012
 Added support for scaffolding in the beta version. Please use the ExpCov parameter to input the expected coverage of a unique contig, this value can be well determined by looking at the "CovHist.txt" file generated by single end assembly. The first column in the file means the coverage, the second column records the number of k-mers with this coverage. The non-noise peak value should be used for ExpCov.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Mon, 23 Jul 2012 16:47:49 -0000

--- v19
+++ v20
@@ -2,8 +2,8 @@
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
-Last update: July 12, 2012
-Fixed basic bugs.
+Last update: July 23, 2012
+Improved speed for the assembly and improved performance of the scaffolder of the beta version.
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1
@@ -21,6 +21,8 @@
 Last update: July 8, 2012
 Added support for scaffolding in the beta version. Please use the ExpCov parameter to input the expected coverage of a unique contig, this value can be well determined by looking at the "CovHist.txt" file generated by single end assembly. The first column in the file means the coverage, the second column records the number of k-mers with this coverage. The non-noise peak value should be used for ExpCov.
 
+
+
 June 20, 2012
 A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, k 31 g 25 memory peak: ~ 15 GB, k 51 g 20, memory peak: ~20GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Thu, 12 Jul 2012 16:41:06 -0000

--- v18
+++ v19
@@ -2,8 +2,8 @@
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
-Last update: July 8, 2012
-Added full support for scaffolding. (Please read below)
+Last update: July 12, 2012
+Fixed basic bugs.
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Sun, 08 Jul 2012 22:57:35 -0000

--- v17
+++ v18
@@ -2,8 +2,8 @@
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
-Last update: July 7, 2012
-Added support for scaffolding. (Please read below)
+Last update: July 8, 2012
+Added full support for scaffolding. (Please read below)
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1
@@ -18,8 +18,8 @@
 
 Updates: 
 
-Last update: July 7, 2012
-Added support for scaffolding in the beta version. Libraries with reasonable errors in insert sizes/orientations are suggested, usually mate pair libraries with insert sizes <5k are good enough. Please use the ExpCov parameter to input the expected coverage of a unique contig, this value can be well determined by looking at the "CovHist.txt" file generated by single end assembly. The first column in the file means the coverage, the second column records the number of k-mers with this coverage. The non-noise peak value should be used for ExpCov.
+Last update: July 8, 2012
+Added support for scaffolding in the beta version. Please use the ExpCov parameter to input the expected coverage of a unique contig, this value can be well determined by looking at the "CovHist.txt" file generated by single end assembly. The first column in the file means the coverage, the second column records the number of k-mers with this coverage. The non-noise peak value should be used for ExpCov.
 
 June 20, 2012
 A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, k 31 g 25 memory peak: ~ 15 GB, k 51 g 20, memory peak: ~20GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Sat, 07 Jul 2012 20:59:26 -0000

--- v16
+++ v17
@@ -3,7 +3,7 @@
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
 Last update: July 7, 2012
-Added support for scaffolding.
+Added support for scaffolding. (Please read below)
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1
@@ -19,7 +19,7 @@
 Updates: 
 
 Last update: July 7, 2012
-Added support for scaffolding.
+Added support for scaffolding in the beta version. Libraries with reasonable errors in insert sizes/orientations are suggested, usually mate pair libraries with insert sizes <5k are good enough. Please use the ExpCov parameter to input the expected coverage of a unique contig, this value can be well determined by looking at the "CovHist.txt" file generated by single end assembly. The first column in the file means the coverage, the second column records the number of k-mers with this coverage. The non-noise peak value should be used for ExpCov.
 
 June 20, 2012
 A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, k 31 g 25 memory peak: ~ 15 GB, k 51 g 20, memory peak: ~20GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Sat, 07 Jul 2012 20:40:35 -0000

--- v15
+++ v16
@@ -2,8 +2,8 @@
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
-Last update: June 23, 2012
-With bugs fixed for the beta version.
+Last update: July 7, 2012
+Added support for scaffolding.
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1
@@ -16,7 +16,12 @@
 
 Results: We demonstrate that the exhaustive decomposition of reads into all overlapping k-mers is unnecessary. Instead, a sparse graph structure, saving only a sparse subset of observed k-mers and the links between them, greatly reduces the memory space requirements for de novo genome assembly. We traverse this sparse k-mer graph to assemble contigs and ultimately complete the genome assembly. We adopt a Dijkstra-like breadth-first search algorithm to circumvent sequencing errors and resolve polymorphisms. Here, we test our SparseAssembler with both simulated and real data, achieving ~90% memory savings and retaining assembly accuracy, but without sacrificing speed in comparison to existing assemblers.
 
-Updates: June 20, 2012
+Updates: 
+
+Last update: July 7, 2012
+Added support for scaffolding.
+
+June 20, 2012
 A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, k 31 g 25 memory peak: ~ 15 GB, k 51 g 20, memory peak: ~20GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Sat, 23 Jun 2012 14:46:59 -0000

--- v14
+++ v15
@@ -2,7 +2,8 @@
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
-Last update: June 22, 2012
+Last update: June 23, 2012
+With bugs fixed for the beta version.
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1
@@ -16,7 +17,7 @@
 Results: We demonstrate that the exhaustive decomposition of reads into all overlapping k-mers is unnecessary. Instead, a sparse graph structure, saving only a sparse subset of observed k-mers and the links between them, greatly reduces the memory space requirements for de novo genome assembly. We traverse this sparse k-mer graph to assemble contigs and ultimately complete the genome assembly. We adopt a Dijkstra-like breadth-first search algorithm to circumvent sequencing errors and resolve polymorphisms. Here, we test our SparseAssembler with both simulated and real data, achieving ~90% memory savings and retaining assembly accuracy, but without sacrificing speed in comparison to existing assemblers.
 
 Updates: June 20, 2012
-A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, (k 31 g 25) memory peak: ~ 23 GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.
+A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, k 31 g 25 memory peak: ~ 15 GB, k 51 g 20, memory peak: ~20GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.
 
 
 If you have suggestions or questions, please contact Chengxi Ye: cxy AT umd.edu.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Fri, 22 Jun 2012 10:30:10 -0000

--- v13
+++ v14
@@ -15,7 +15,7 @@
 
 Results: We demonstrate that the exhaustive decomposition of reads into all overlapping k-mers is unnecessary. Instead, a sparse graph structure, saving only a sparse subset of observed k-mers and the links between them, greatly reduces the memory space requirements for de novo genome assembly. We traverse this sparse k-mer graph to assemble contigs and ultimately complete the genome assembly. We adopt a Dijkstra-like breadth-first search algorithm to circumvent sequencing errors and resolve polymorphisms. Here, we test our SparseAssembler with both simulated and real data, achieving ~90% memory savings and retaining assembly accuracy, but without sacrificing speed in comparison to existing assemblers.
 
-Last update: June 20, 2012
+Updates: June 20, 2012
 A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, (k 31 g 25) memory peak: ~ 23 GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Fri, 22 Jun 2012 10:29:40 -0000

--- v12
+++ v13
@@ -1,8 +1,8 @@
 [[project_admins]]
 SparseAssembler: Sparse k-mer Graph for Memory Efficient de novo Genome Assembly
-
 Chengxi Ye, Zhanshan (Sam) Ma, Charles H. Cannon, Mihai Pop, Douglas W. Yu
 
+Last update: June 22, 2012
 
 The formal version of our work, titled 'Exploiting Sparseness in de novo Genome Assembly' has been published in BMC Bioinformatics.
 http://www.biomedcentral.com/1471-2105/13/S6/S1

WikiPage Home modified by Chengxi Ye

Chengxi Ye — Wed, 20 Jun 2012 04:16:55 -0000

--- v11
+++ v12
@@ -15,7 +15,7 @@
 
 Results: We demonstrate that the exhaustive decomposition of reads into all overlapping k-mers is unnecessary. Instead, a sparse graph structure, saving only a sparse subset of observed k-mers and the links between them, greatly reduces the memory space requirements for de novo genome assembly. We traverse this sparse k-mer graph to assemble contigs and ultimately complete the genome assembly. We adopt a Dijkstra-like breadth-first search algorithm to circumvent sequencing errors and resolve polymorphisms. Here, we test our SparseAssembler with both simulated and real data, achieving ~90% memory savings and retaining assembly accuracy, but without sacrificing speed in comparison to existing assemblers.
 
-Last update: June 17, 2012
+Last update: June 20, 2012
 A beta version which is more efficient and contains error correction is released. Some preliminary results: NA12878 human genome, (k 31 g 25) memory peak: ~ 23 GB. ERA000206 E. coli, ~600X, (k 51 g 25) memory peak: ~ 120 MB.