ZOOM Frequently Asked Questions
Question: Can I put reads of different lengths in the same file?
Question: Is the input read data case sensitive?
Question: Can I get all mapped positions for each read, besides the uniquely mapped information?
Question: In which cases can ZOOM achieve 100% sensitivity?
Question: How to get better sensitivity using 100% sensitivity seeds without too much running time?
Question: Can ZOOM find short indels?
Question: The quality of 3'-end reads is not so good, what should I do?
Question: How many reads can ZOOM deal with in 8G RAM?
Question: What should I do if the read number is too large to fit in the RAM?
Question: Can ZOOM utilize the quality score of reads to enhance mapping results?
Question: Can ZOOM output the structural variation according to the output of paired-end mapping?
Question: Can ZOOM output the SNP candidates or the INDEL candidates?
Question: Are there restrictions on the length of reads label?
Question: Can ZOOM deal with 454 data or Helico data?
Question: Can I put reads of different lengths in the same file?
Answer. Yes, ZOOM will automatically call different parameter sets for different read lengths and the results will be merged.
Question: Is the input read data case sensitive?
Answer. No, "a"= "A", "c"="C", "g"="G", "t"="T"="u"="U", and all other letters are "N". If you have different requirements, please contact us.
Question: Can I get all mapped positions for each read, besides the uniquely mapped information?
Answer. Yes, use the parameter <-mk> to output top N best mapping results for each read. Set N very large if you want all mapping results for each read.
Question: In which cases can ZOOM achieve 100% sensitivity?
Answer. ZOOM designs a framework to construct the efficient spaced seeds sets which can achieve 100% sensitivity for a large range of read lengths and mismatch numbers. All cases in this release are listed in the section entitled "100% sensitivity cases". For cases with more mismatch numbers and cases with insertion/deletion, ZOOM also has good sensitivity. If you do need 100% sensitivity beyond the listed cases, please contact us.
Question: How do I get better sensitivity using 100% sensitivity seeds without using a lot of time?
Answer. Set the "-mm" to be higher without "-sv" parameter first. Then extract those unmapped reads to run ZOOM using 100% sensitivity seeds with "-sv" parameter.
Question: Can ZOOM find short indels?
Answer. Yes. However, ZOOM can only find one gap with any length. The speed is about five times slower than the mode only allowing mismatches when each indel is allowed.
Question: The quality of 3'-end reads is not so good, what should I do?
Answer. You can set a threshold between high quality bases and low quality bases using the "-rh" parameter, then ZOOM will neglect those low quality bases when mapping.
Question: How many reads can ZOOM deal with in 8G RAM?
Answer.For command-line version, we suggest 25-30 million reads for 8G RAM. If you double your RAM, you can also double the data size. For the GUI verison, ZOOM can split reads into small pieces. You can modify the size of the small pieces to run on a different size of RAM. Please refer to Section 2.6.
Question: What should I do if the read number is too large to fit in the RAM?
Answer. Please split the read file into several parts which could be fit in the RAM. If you have 8G RAM, 25 million reads in one file are recommended. If your memory doubles, then the number of reads in one file could double too. Run the several parts separately, and merge the mapping results into one result file. If you want to do assembly too, please use the option "-mii" to load in the merged result file to do assembly. If the data set is paired-end reads, please mate up the reads in one file and carry out the above operations.
Question: Can ZOOM utilize the quality score of reads to enhance mapping results?
Answer. Yes. For Illumina/Solexa data, ZOOM adopts two ways to utilize quality score to enhance mapping results. The first way is to only count mismatches occurring on high quality positions. The second one is to utilize quality score of each base of read to compute the mapping probability of possible alignments for each read and choose best or top N mapping results according to the mapping probability. For ABI SOLiD data, because ZOOM can differentiate possible genomic differences from sequencing errors on color space, ZOOM computes the mapping probability of alignments for each read utilizing both quality scores and the probability of a SNP occurring in the organism you sequenced. Then it uses the mapping probability to assess and choose best or top N mapping results for each read.
Question: Can ZOOM output the structural variation according to the output of paired-end mapping?
Answer. Not yet. In version 1.3, ZOOM outputs read pairs mapped in the distance range. You can judge whether there is stuctural variation by the mapping offsets and direction of the two mates of one pair. ZOOM offers the "process unmapped reads" function, which will map those unmapped reads with a different distance range or map them in single-end mode. This might help you to identify structural variation.
Question: Can ZOOM output the SNP candidates or the INDEL candidates?
Answer. Yes. You can ask ZOOM to carry out the post-analysis to find SNP candidates and view them in an intuitive way.
Question: Would it be possible to have a parameter for % mismatch or insertion/deletions instead of a fixed number?
Answer. Yes. ZOOM use the "-mmpct" parameter to denote the percent of mismatches. The "-edpct" parameter and the "-idpct" parameter denote the percent of edit distance and the percent of insertion/deletion length respectively. When using percent parameter on the data set with reads of various length, ZOOM will handle reads of different lenght separately (with ¨Cngrpsv parameter), which is slower than the case a fixed number is assigned where all reads are handled together.
Question: Are there restrictions on the length of reads label?
Answer. No. We do advise against the use of spaces inside the label for the 'one read per line' format, because spaces aids in identifying where the read data field begins.
Question: Can ZOOM deal with 454 data or Helico data?
Answer. ZOOM is optimized for Illumina/Solexa and ABI SOLiD data. ZOOM can get good mapping results on these two instruments. However, the sequencing error types of 454 instrument and Helico instrument are quite different, which contain much short indels. ZOOM can't guarantee good mapping results because currently ZOOM can only handle one gap of any length rather than many gaps. However, you could give it a try since ZOOM can handle reads over 200bp and can deal with reads of variable lengths automatically. Any feedback would be appreciated. We look forward to supporting these two instruments in the future.
Question: Can ZOOM schedule multiple jobs on multiple CPUs of one server or multiple servers?
Answer. Yes. Please configure the server address using the Configuration button, and ZOOM will split the job into several tasks, schedule among these servers and collect the mapping results automatically. You can also choose the data size of each task running on each CPU according to the RAM of your server.
