o f?@s`ddlmZddlZddlZddlZddZddZdd Zd d Z d d Z ddZ ddZ dS))data_ioNc Cs<|d}t|d|}t||||\}}t|||dSN)r load_peaksZload_regions_from_bwwrite_regions_npz) peaks_pathZfa_pathZbw_pathsout_path region_width half_widthpeaks_df sequencescontribsr8/users/shouvikm/ChromBPNet/finemo_gpu/src/finemo/main.pyextract_regions_bw srcC*|d}t||\}}t|||dSr)rZload_regions_from_chrombpnet_h5rZh5_pathsr r r r rrrrextract_regions_chrombpnet_h5rcCrr)rZload_regions_from_bpnet_h5rrrrrextract_regions_bpnet_h5rrcCs,|d}t|||\}}t|||dSr)rZload_regions_from_modisco_fmtr)Z shaps_pathsZohe_pathr r r r rrrrextract_regions_modisco_fmt"src$Cst}ddlm}t|\}}|jd}|ddkr#td|d|d}|durSt|||}|j}||jdksC||jdkrRtd|jd|jd |n|jd}|d krad }d }n|d krjd }d}n|dkrsd}d }n|dkr{d}d}t |||\}}|jd}|jd}| |||||||| | | | | | \}}t |} t |j dd}!tj|ddtj|d}"|durt| |||!||t|!||"nt| ||!||t|!|"|||||dO}tj|d}#t||#dS)Nr) hitcallerrrzRegion width of z is not divisible by 2.zInput sequences of shape z% and/or input contributions of shape z) are not compatible with region count of ppcwmFphThpZhcwmhhZpeak_id)nameexist_okz peaks_qc.tsv)r num_regionsZuntrimmed_motif_width num_motifszparameters.json)localsrrload_regions_npzshape ValueErrorrZheightload_modisco_motifsZ fit_contribsplZ DataFrameZwith_row_countosmakedirspathjoinZ write_hitsZwrite_qcZwrite_hits_no_peaksZwrite_qc_no_peaksZ write_params)$ regions_pathrmodisco_h5_pathZchrom_order_pathout_dircwm_trim_thresholdalpha step_size_max step_size_minconvergence_tol max_steps batch_size step_adjustdevicemodeno_post_filterparamsrr rr r r r!Z motif_typeZuse_hypothetical_contribs motifs_dfcwmsr" motif_widthhitsZqchits_dfZqc_dfZ out_path_qcZout_path_paramsrrr call_hits*sj         rBc Csddlm}t|\}}t|jdkr||} nt|jdkr-|dddddf|} | jdd} |d} t|d| } tj|dd} tj|| | | dd}t |dd\}}| t d d k d }|jd}|| |\}}|| | | ||| |\}}}tj|dd tj|d }t||t|||||||tj|d}||||tj|d}|||tj|d}|||tj|d}||||dS)Nr) evaluationrT)ZlazyrrZ motif_strand+Z motif_namerzmotif_occurrences.tsvzmotif_cooocurrence.pngZCWMszhit_vs_seqlet_counts.pngz report.html)r$rCrr%lenr&rZ load_hitsZload_modisco_seqletsr(filterr)ZcolZ get_columnZto_numpyZget_motif_occurencesZ seqlet_recallr*r+r,r-Z write_occ_dfZwrite_recall_dataZplot_hit_distributionsZ!plot_peak_motif_indicator_heatmapZ plot_cwmsZplot_hit_vs_seqlet_countsZ write_report)r.Z hits_pathr/rr0modisco_region_widthrCr rregionsr Zmodisco_half_widthr rAZ seqlets_dfr=Z cwms_modiscoZ motif_namesr?Zocc_dfZcooocZ recall_dataZ recall_dfr>Zocc_pathZ coooc_pathZplot_dirZ plot_pathZ report_pathrrrreportms>        rJc Cst}|jddd}|jdtjdd}|jddtd hd d d |jd dtddd|jddtddd|jddtddd|jddtddd|jddtddd|jddtd d!d|jd"d#td$d%d|jd&d'd(d)d*|jd+d,td-d.d|jd/d0td1d2d|jd3d4td5d6d|jd7d8td9d:d|jd;dd|jd?d@tdAdBd|jdCdDtdEdFd|jdGtjdHd}|jddtddId|jd&dJtddKd|jd?dLtddMdNdO|jddPtddQd|jdRdStdTdUd|jdVtjdWd}|jd7dXtddMdYdO|jddPtddQd|jdRdStdTdUd|jdZtjd[d}|jd7dXtddMdYdO|jddPtddQd|jdRdStdTdUd|jd\tjd]d}|jd7dXtddMdYdO|jddPtddQd|jdRdStdTdUd|jd^tjd_d}|jd+d`tddad|jd"dbtddMdcdO|jddPtddQd|jdRdStdTdUd|jddtjded}|jd dtddfd|jdgdhtddid|jddtddjd|jddtddkd|jddtddd|jdldmtdndod| } | j dkrt | j | j | j| j| j| j| j| j| j| j| j| j| j| j| j| jdS| j dGkr%t| j | j| j| j| j dS| j dVkr6t!| j"| j| j dS| j dZkrKt#dpt!| j"| j| j dS| j d\kr\t$| j"| j| j dS| j d^krot%| j&| j'| j| j dS| j ddkrt(| j | j)| j| j | j| j*dSdS)qNTcmd)requiredZdestz call-hitsz@Call hits on provided sequences, contributions, and motif CWM's.)Zformatter_classhelpz-Mz--moder>rrrrzThe type of attributions to use for CWM's and input contribution scores, respectively. 'h' for hypothetical and 'p' for projected.)typedefaultZchoicesrMz-rz --regionszpA .npz file of input sequences and contributions. Can be generated using `finemo extract-regions-*` subcommands.)rNrLrMz-mz --modisco-h5z2A tfmodisco-lite output H5 file of motif patterns.z-pz--peakszA peak regions file in ENCODE NarrowPeak format, exactly matching the regions specified in `--regions`. If omitted, outputs will lack absolute genomic coordinates.)rNrOrMz-Cz --chrom-orderzA tab-delimited file with chromosome names in the first column to define sort order of chromosomes. Missing chromosomes are ordered as they appear in -p/--peaks.z-oz --out-dirz!The path to the output directory.z-tz--cwm-trim-thresholdg333333?zNThe threshold to determine motif start and end positions within the full CWMs.z-az--alphag333333?zThe L1 regularization weight.z-fz--no-post-filterZ store_truezDo not perform post-hit-calling filtering. By default, hits are filtered based on a minimum correlation of `alpha` with the input contributions.)actionrMz-sz--step-size-maxg@z The maximum optimizer step size.z-iz--step-size-ming{Gz?z The minimum optimizer step size.z-Az --step-adjustgffffff?zThe optimizer step size adjustment factor. If the optimizer diverges, the step size is multiplicatively adjusted by this factorz-cz--convergence-tolgMb@?zoThe tolerance for determining convergence. The optimizer exits when the duality gap is less than the tolerance.z-Sz --max-stepsi'z)The maximum number of optimization steps.z-bz --batch-sizeiz%The batch size used for optimization.z-dz--deviceZcudazBThe pytorch device name to use. Set to `cpu` to run without a GPU.zextract-regions-bwz@Extract sequences and contributions from FASTA and bigwig files.z0A peak regions file in ENCODE NarrowPeak format.z--fastazcA genome FASTA file. If an .fai index file doesn't exist in the same directory, it will be created.z --bigwigsrEzvOne or more bigwig files of contribution scores, with paths delimited by whitespace. Scores are averaged across files.)rNrLZnargsrMz --out-pathz!The path to the output .npz file.z-wz--region-widthiz?The width of the input region centered around each peak summit.zextract-regions-chrombpnet-h5zKExtract sequences and contributions from ChromBPNet contributions H5 files.z--h5szrOne or more H5 files of contribution scores, with paths delimited by whitespace. Scores are averaged across files.zextract-regions-h5zExtract sequences and contributions from ChromBPNet contributions H5 files. DEPRECATED: Use `extract-regions-chrombpnet-h5` instead.zextract-regions-bpnet-h5zFExtract sequences and contributions from BPNet contributions H5 files.zextract-regions-modisco-fmtzDExtract sequences and contributions from tfmodisco-lite input files.z --sequencesz9A .npy or .npz file containing one-hot encoded sequences.z--attributionszOne or more .npy or .npz files of hypothetical contribution scores, with paths delimited by whitespace. Scores are averaged across files.rJzs  C .