This script calculates the probability of finding the top variant in a library of sequences and the probability of finding the top K variants as Nov published it in 2012. The results from his paper were reproduced and can be found as a script in the examples folder. This script builds on top, with the ability to mix different proportions of NNS, DKS or NNN strategies. In the example, 7 DKS + 1 NNS position and 7 NNS positions were applied. A computer with a good GPU is recommended to handle the load.
Install requirements.txt using python>3.0
python3 top_variants.py [OPTIONS]--output_dir: Output directory (default: "output")--get_top_k_variants: Get top k variants--probability_T1: Probability of finding the top variant (default: 0.8371929152161488)--ks_list: List of k values concatenated by commas, e.g., "1,10,100,1000" (default: "1,10,100,1000")--degeneracy_strategy_list: List of degeneracy strategies concatenated by commas, e.g., "NNN,NNS,DKS" (default: "NNN")--counts_list: List of counts concatenated by commas, e.g., "1,2,3" (default: "3")--library_size: Library size (default: 79041)--chunk_size: Chunk size to generate the tensor probabilities (default: 1000)--print_interval: Number interval to print the iteration step (default: 100000)--log_output: Log output--restart_from_iteration: Restart from iteration
python3 ../top_variants.py --output_dir="output" --degeneracy_strategy_list="NNN" --counts_list="3" --library_size=175835 --print_interval=1000 --log_output
python3 top_variants.py --output_dir="output" --get_top_k_variants --probability_T1=0.95 --ks_list="1,10,100,1000"