Table 1. Error filters used in the computational pipeline.

Filter name	Definition
Improper alignment	We rejected reads with less than 30-bp alignment or more than 3 mismatches to the reference genome.
Diagnostic motif	We rejected reads without L1Hs diagnostic G motif (position 6012 relative to the L1Hs Repbase consensus) [29].
Chimera within L1 segment	We rejected reads with less than 95% identity (> 4 mismatches) to the L1Hs 3' end consensus sequence.
Chimera within poly-A tail	We rejected reads at risk of being chimeric [12]. We applied BLAST to find the best alignments for retrotransposon and non-retrotransposon segments from hg19. Reads were removed as a putative chimera when the sequences of two segments overlapped > 10 bp with A% (percentage of adenine nucleotides) ≥ 50% or overlapped 6–10 bp with A% < 50%.
Subfamily filter	We rejected putative somatic insertion sites that overlapped with L1 young subfamilies (L1Hs and L1PA2–4) reference insertions.
Known non-reference filter	We rejected putative somatic insertion sites that overlapped with known non-reference L1 insertions in euL1db [30].
Misaligned reads	We rejected reads at risk of being misaligned, defined as inconsistent BWA and BLAT alignment.
Local structural variation (SV)	We rejected reads at risk of being derived from a nearby reference L1Hs [12]. We extracted 2 kb downstream of aligned non-retrotransposon segment from hg19 and aligned the full contig against this sequence by BLAT to exclude potential genomic rearrangement events.
Observed in common	We rejected putative somatic insertion sites observed in two or more individuals.
PCR duplicate	We rejected somatic insertion sites without supporting PCR duplicates.