Evaluating a Read Mapper on Simulated Data
When evaluating read mappers, we should always focus on well defined sets of reads:
- Reads with no polymorphisms.
- Reads with 1, 2, 3 or more SNPs.
- Reads with specific types of short indels (<10bp).
- Reads with larger structural variants (>100bp).
SNPs and errors are different because SNPs can lead to mismatches in high-quality bases. In addition to integrating according to the metrics above, we could separate results by the number of errors in each read.
Should also be grouped according to whether reads are paired-end or single-end and according to read-length.
- Speed (millions of reads per hour)
- Memory requirements
- Size of output files
- Raw count of mapped reads
The key quantities are:
- How many reads were not mapped at all?
- How many reads were mapped incorrectly? This is the least desirable outcome.
- How many reads were mapped correctly?
Correct mapping should be defined as:
- Most stringent: matches simulated location and CIGAR string.
- Less stringent: overlaps simulated location at base-pair level, CIGAR string and end positions may differ.
- Incorrect: Doesn't overlap simulated location.
We should evaluate mapping qualities by counting how many reads are assigned each mapping quality (or greater) and among those how many map correctly or incorrectly. This gives a Heng Li graph, where one plots number of correctly mapped reads vs. number of mismapped reads.
Available Test Datasets
- Location: wonderland:~zhanxw/BigSimulation
no polymorphism ; 1, 2, 3 SNP ; Deletion 5, 30, 200; Insertion 5, 30
- Quality String
Picked the 75 percentile of Sanger Iluumina 108 mer test data set
both base space and color space both single end and paired end, and paired end reads are given insert size 1500.
- Program (generator)
generator [bs|cs] [se|pe] [exact|snpXX|indelXX|delXX] -n numbers -l readLength -i insertSize exact: Accurate sample from reference genome snpXX: Bring total XXX SNP for a single read or a pair of reads indelXX: Insert a random XX-length piece for a single read, or at the same position for a paired reads delXX: Delete a random XX-length piece for a single read, or at the same position for a paired reads e.g. ./generator bs se exact -n 100 -l 35
Simulation file are named like: BS_SE_EXACT_1000000_35, meaning base space, single end, exact (no polymorphism), 1M reads, 35 bp per read. For each read, the tag was named in a similar way to Sanger's.
Bulk statistics result
Running time (all submitted to the MOSIX client node)
BWA(second) Karma(second) Scenarios 2594 7182 BS_SE_DEL200_1000000_50.fastq 2641 -1 BS_SE_DEL30_1000000_50.fastq 2355 -1 BS_SE_DEL5_1000000_50.fastq 441 7941 BS_SE_EXACT_1000000_50.fastq 809 282 BS_SE_INDEL30_1000000_50.fastq 2217 -1 BS_SE_INDEL5_1000000_50.fastq 645 7206 BS_SE_SNP1_1000000_50.fastq 1102 -1 BS_SE_SNP2_1000000_50.fastq 1142 -1 BS_SE_SNP3_1000000_50.fastq 6536 8874 BS_PE_DEL200_1000000_50_?.fastq 6699 9017 BS_PE_DEL30_1000000_50_?.fastq 6468 9033 BS_PE_DEL5_1000000_50_?.fastq 1743 10112 BS_PE_EXACT_1000000_50_?.fastq 2305 231 BS_PE_INDEL30_1000000_50_?.fastq 5703 2989 BS_PE_INDEL5_1000000_50_?.fastq 1974 3718 BS_PE_SNP1_1000000_50_?.fastq 2396 3339 BS_PE_SNP2_1000000_50_?.fastq 2817 3131 BS_PE_SNP3_1000000_50_?.fastq 4362 16074 CS_PE_DEL200_1000000_50_?.fastq 4385 -1 CS_PE_DEL30_1000000_50_?.fastq 4373 9287 CS_PE_DEL5_1000000_50_?.fastq 773 -1 CS_PE_EXACT_1000000_50_?.fastq 1735 3142 CS_PE_INDEL30_1000000_50_?.fastq 4023 8591 CS_PE_INDEL5_1000000_50_?.fastq 1034 10528 CS_PE_SNP1_1000000_50_?.fastq 2236 -1 CS_PE_SNP2_1000000_50_?.fastq 3810 6617 CS_PE_SNP3_1000000_50_?.fastq 7129 1493 CS_SE_DEL200_1000000_50.fastq 7115 1513 CS_SE_DEL30_1000000_50.fastq 7065 1542 CS_SE_DEL5_1000000_50.fastq 1544 1666 CS_SE_EXACT_1000000_50.fastq 2954 289 CS_SE_INDEL30_1000000_50.fastq 6547 1390 CS_SE_INDEL5_1000000_50.fastq 1690 1661 CS_SE_SNP1_1000000_50.fastq 2853 1449 CS_SE_SNP2_1000000_50.fastq 4039 1237 CS_SE_SNP3_1000000_50.fastq