String of 'N's at same location in Fasta file

Question

[zdiazmar]

Hi all,

I am relatively new to bioinformatics and using genomic data. I have generated some ddRAD data using the Petersen et al. 2012 protocol and STACKS to call SNPs (default parameters for -m, -M, and -n). I have filtered sequences to retain 'high-quality' reads. My average read depth coverage per locus is 29X and the average amount of missing data per locus is 7%. Taking a look at my Fasta file I am noticing that for most sequences there is a string of ~10 'N's at around the same location. Here is a subset of sequences.

Locus1: AATTCTTGCAGTGAGAAGTACTGGTGAGTTTCATCCTCATGTTCTGTTCCTATAGTACTGTGGTATACTTATTTTGGATCTTCTTATATGATTAATGGATCNNNNNNNNNNTATTACATACTATTCTGATATCTGTCTGTCGAAATTACAGGCATTTCTGATGAGAAGGTGCATAATCATGAACTCAACCTTAAGGAAGCCGCTTCCACCGG

Locus2: AATTCCTTCGAGTATTGAAGGGATGGCATCTTCATCACCATCATAGAATTTCTGAATCCTCTTCTTCAGCTCTATACGAAAAGCAAAAATCAGTGCAATGNNNNNNNNNNAGTCGAGATTCAACTCCCCAAACAGCCTCATAAGAGAACTAGTAATAAACATTACTTACTGAACTACACTATCGGAAGATTTGTGGCACCGCATAAACCGGACCGTATTTTTCAAAATGGACTTTGACTTATACCGAACCG

Locus3: AATTCAACAACAATTAACGTAGTACTCCATCAAGGTTCCAATCAAAACATCTCTTCTATGTCACTCCATAATAATATAACCTCACTATGCATATCCACTANNNNNNNNNNTCCACCTTGTTATTGCTTGATCCGATTCGGGTCATGAAGAACGGGTTCGGGTCGATATTGATGGCCACTTCTGGTGCTACTTTCGCGGAATGCGGGCCCG

Locus4: AATTCACCTCATTCTTGCGGGTGTTGGGGAGCTATCCTATGGATATGACCCCGTGGTGTCCTTCTAGAGGAGACTAGTAATTAATTAATTTAAAAGTAAANNNNNNNNNNTGAAACTGTGTTTCTTTATTCTTGATTTCACGTTCTTTACCTAAAATACCTACCTACCTGAACTTCTTTCCTCGTGCGGGATTCGAATTGGTGGTGGCCGGAACCAGTACCATAGTCGGTCTGATTCAGTCAAAGTTTCAAGATGTACGCGAACCCGATTTAACCTCCCG

Locus5: AATTCTGATAATCCATTTGTTTGCTCATCTAGTTCTTATGATACAATTATCTGCATCTTTTTCCTTTATACCCGCCGACTTGTTTTCTGCACTAGTAGTGNNNNNNNNNNTAGGTTGGCCATACCCGCACATGCCATAACGACAACGAACATGCCTTGGACATCGATTGAAGCATTTCCCACAATGTCTTCGGTTTCGATTGATGTTCCGGGGGGGCCCTATGAAAATTTGGAAATACGATGCTTTGATAGAGAAAAGGATCCG

Locus6: AATTCAACATAGTTCATGAATCGGGTTATCTATTTTTTACCTGCATGTACCTGGCCAGAACTAAAAAGTCGGTTTCTTGAACCAACTCCAATATCTCTCCTNNNNNNNNNNGGGTCGAGAAGAGGGACTCGGACTACGGATGTCAGGGCATTGAACCTCTTTGCAAGATTAAGGAAGTCCTCATTCATTTGCTCTCGTGGCCATCTAGCCGGTTCGACCCATTTAGTATTCTTGGTCCGACCCGACCCGAATAGGTCGGACCCGATTGAGAAAAACCG

Locus7: AATTCACTGAATGTGTCACGGACTAATAGCATCTATCCAATGATTAGAGGAAATTATTTTAGTTTTTTGGGCAGTGGAAAACTAAAAAATATGGTTTAAANNNNNNNNNNGAAAAGGGGAAAATCTAAATACGAACTTAACGAAAACCCAATACTCTGACAAGGATACCCAATAACCACACATTGTAAAGCAAAAACATGAAGTCAACCGG

Locus8: AATTCTGACAGGAATCATAATGGGATGTGCGCTTTTATTCATGACTCCGTTATTTGAGTATATACCATTGGTATGCTGCAGGCATTTATTTCTTTGATATNNNNNNNNNNAAATTACTTTCCCCCATAATTACGTGTGAGGTGATGCTTACCTTCTAAGACCATAAATAAAATCATCCAAAACCCCTCAACCTGGGATTGGCTGTGGCCGGTCCGGTTCGGTTTTCATTTAAAAATCTGTTCTGTAAATTTCTGTCCGGTCCG

Locus9: AATTCAATTATTGATAGGTTCCAGTAAACTGTATTATTAGTAAGCTAACAGAAGCAGTTGGCGTCAAGATCCCATGAAATAAGTTAGAAGAATCATCATTNNNNNNNNNNTTAGTTCTAGCTACGAATGAATGGAAAAGAGCAGATGGATCAAGAAAATTAAACATTTCCTGGGAAATCCCCATCTGTTAATAGGAGGAGGAGAAGGTTACCGGAGGGAGCTGATTATTCATCTCCTTCTCACCTTCCGTGAATAGCCG

Locus10: AATTCAGAAAAGGAGAGGGACAAATGCTGAAATCCAAACCTCAAGTCCCACAAAAGTGATTGACCATTACACTGGAGATGGCTCTCCCAAGATGACGTCTNNNNNNNNNNAATGGATATGTTAAACATAGACCAAATAACTATAACCTCACAAAGAAATGTGTACATTATGTAGATCTTTCATGAACAAAAAGCAAAATAATACAGCCCGGAGACCGAAGCTCCG

Locus11: AATTCTTTTCACCACCCACAAACCATACCTTGATTTGTTGATTCAACTTGCAGGTATCTATGTTGGAGGAAGCAAGATTGTTCATTTCAGACCTGACCCANNNNNNNNNNGATTCGAGTACGGGGTCCGACCTTCACTCTTCCTAGCCAAAGTCAGAGGCGGCACATGCACCACCGCACCCTCTGACCCGCCCGAAACAGTCATCGACCGGTGGTATTTCGTTATAGTTTCCG

Locus12: AATTCATGATCGGTTCCTTTTTAAGTCACTTCTTATTCACATCATGTACAAAATGAGACCAGACCGATTGATCCGAGTGCCCATACAAAAGACGATTAATNNNNNNNNNNATACCAACTACTAACCTGTAAGATTGGTTCCATTGGGGATGCTCACCGTAGAATTGAGAAACCATGAGCAAACTTTCCGACGTCGGATTGTCGACAACCGG

Does anyone have any ideas of why this might be happening? Thank you for any input!

Answer 1

[zdiazmar]

In case anyone out there also runs into this, it seems that this is due to non-overlapping paired-end reads. Here is the response from a collaborating bioinformatician:

Stacks added the Ns as part of its analysis process. Here's a description (from the Stacks manual) of how Stacks handles paired-end data:

Stacks directly supports paired-end reads, for both single and double digest protocols. For double-digest RAD, both the paired-end reads are anchored by a contig and Stacks will assemble them into two loci. In both cases, the paired-end contig/locus will be merged with the single-end locus. If the loci do not overlap, they will be merged with a small buffer of Ns in between them.

So the Ns indicate that the paired-end reads at a locus do not overlap, and the actual number of Ns is unknown, although if you know the length distribution of the library (e.g. from a BioAnalyzer trace) you'll have a rough idea of how many Ns there could be.