Fixed a bug that few split-reads are marked in the wrong direction, t…

…his leads to more accurate determination of the insertion strand and breakpoint.

Bugs fixed.\n1. inaccurate insertion strand annotation in germline in…

…sertion detection mode.\n2. inaccurate insertion strand annotation in ALU mode

update dependency of python2 to python3

fixed bugs where default awk is not working properly. Now specificall…

…y use gawk instead of default awk.

fixed the issue that cat has a maximum file limitation

added a parameter -k, to enable output those singleton insertions wit…

…h FDR equals to 0 for debugging

Add -G for users to input number of genomes used in the library and T…

…EMP2 can automatically choose threshold to detect potential de novo insertions.

Change -C from self-defined read cutoff to self-defined frequency cut…

…off.

v0.1.2

1. By default, TEMP2 hypothesize that de novo insertions that are derived during individual development is full-length. However, a few transposons does have truncated insertions, such as 5' truncated L1 elements. Now, users can use the option -T to allow truncated de novo insertions. In this option, TEMP2 will firstly mark those truncated insertions that are supported by at least 3 reads at both ends. Then marked truncated insertions together with full-length insertions are used by TEMP2 for estimating de novo insertion numbers.
2. Usually, Illumina DNA sequencing or other short-read DNA sequencing takes about or more than 10,000 genomes during library construction. Considering the seqeuncing depth is much lower than 10,000X, de novo insertions which only occurs in one or few genomes are only supported by one read. Therefore, TEMP2 regardsingleton reads as potential de novo insertions by default. However, if a special library with limited number of genomes (for example: 50) is used during library constrcution, de novo insertions can be supported by more than one read, known as 2p or 1p1 insertions. In this situation, users can add a self-defined supporting-read cutoff using option "-C cutoff" to determine if a insertion is potential de novo insertion. For example, if 50 genomes are used in library construction and the seqeuncing depth is 50X (100bp paired reads, 400bp fragmnent length), the number of supporting reads for a de novo insertion should be around (fragment_length/read_length/2*depth*2/number_of_genome=400/100/2*50*2/50=4), and you can set -C 7 for a more accurate de novo insertion estimation.

fixed warning bug