1. Alternative Computational Analysis Shows No Evidence for Nucleosome Enrichment at Repetitive Sequences in Mammalian Spermatozoa 
Hélène Royo, Michael Beda Stadler, Antoine Hendrik Felix Marie Peters 
Developmental Cell 
Volume 37, Issue 1, Pages (April 2016)
DOI: /j.devcel
Copyright © 2016 Elsevier Inc. Terms and Conditions

2. Figure 1
Samans' Analysis Pipeline Leads to an Artifactual Signal in Repetitive DNA Elements
(A) Cartoon illustrating the artifactual signal amplification of repeat elements by Samans' pipeline. Sequence reads with unique and multiple mapping positions that originate from repetitive DNA elements are shown in orange and red, respectively, non-repeat reads in light and dark blue. Only multi-mapping reads are amplified in the reported alignments (5- and 3-fold in this arbitrary example). Combined amplification factors affecting all reads in the repeat and non-repeat classes are shown (3 and 2 in this arbitrary example). SNP, single nucleotide polymorphism; BS, binding sites. The gray spheres represent nucleosomes.
(B) The percentage of reads and alignments in the four read classes, as explained in (A), are shown for the bovine and human sperm datasets from Samans et al. (2014). The numbers refer to the amplification factors of reported alignments over the number of actual reads. The higher amplification factors in the bovine dataset may be due to the shorter read length (single-read 84-mers for bovine versus paired-end with an average fragment length of 135 base pairs for human) and the repeat landscape of that genome.
Developmental Cell  , DOI: ( /j.devcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

3. Figure 2
Samans' Pipeline Creates Similar Artifactual Enrichments in Sperm MNase and Control Datasets
(A–C) Alignment densities in three genomic regions (corresponding to the regions displayed in Figure 2A of the Samans et al. (2014) paper) are shown for the sperm MNase sample, a control sample with matched read length and coverage obtained from ENCODE (MCF-7 cells), and a simulated dataset (see Experimental Procedures). Each sample was aligned using the “1 alignment/read” and Samans' pipelines, with both settings tolerating no mismatches. Repetitive DNA elements are shown in shades of blue, with darker blue indicating a higher similarity to the repeat family consensus sequence.
See also Figure S1 and Tables S1 and S2.
Developmental Cell  , DOI: ( /j.devcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions

4. Figure 3
Artifactual Enrichments Variably Affect Different Repeat Families and Have an Inverse Relationship with Sequence Divergence
(A) For each repeat family, the number of reads in the sperm MNase dataset mapping to that family was counted and plotted as log2 reads per million (RPM) for both the “1 alignment/read” and Samans' pipeline. SINE, LINE-1, and Satellite families (including centromeric satellite repeats) are highlighted by colors, and the dotted line corresponds to y = x.
(B) Same as in (A) for individual types of LINE-1 elements. The color indicates the median divergence from the consensus sequence, as measured by the number of nucleotide mismatches per kilobase (substitutions/kb).
(C) Artifactual alignment amplification was quantified for individual types of LINE-1 elements by calculating the ratio of counts obtained using Samans' pipeline over the “1 alignment/read” pipeline. The dashed line corresponds to a linear fit of the sequence divergence (substitutions/kb) to the amplification factor in log2 space, including only repeat types with a median divergence of less than 100 substitutions/kb.
See also Figures S2 and S3.
Developmental Cell  , DOI: ( /j.devcel )
Copyright © 2016 Elsevier Inc. Terms and Conditions