DUfJddlZddlZddlZddlZddlmZddlm Z m Z ddl m Z m Z ddlZddZ ddZd Zd Z ddZdddddddefdZ ddZdS)N)numutils)is_compatible_viewframeis_cooler_balanced)make_cooler_viewalign_track_with_coolerc$g}|D]}tj|tj|z}||dkr4tj||||d}n |dkr3tj||||d}n|dkrhtj||||d}tj||z|ztj||z}ne|dkr=tj ||||tj ||z}n"td || |tt|D](}tj||||z||<)|7tjtj| } || || }}||fS)a@ Flip `eigvecs` to achieve a positive correlation with `phasing_track`. Parameters ---------- sort_metric : str If provided, re-sort `eigenvecs` and `eigvals` in the order of decreasing correlation between phasing_track and eigenvector, using the specified measure of correlation. Possible values: 'pearsonr' - sort by decreasing Pearson correlation. 'var_explained' - sort by decreasing absolute amount of variation in `eigvecs` explained by `phasing_track` (i.e. R^2 * var(eigvec)). 'MAD_explained' - sort by decreasing absolute amount of Median Absolute Deviation from the median of `eigvecs` explained by `phasing_track` (i.e. COMED(eigvec, phasing_track) * MAD(eigvec)). 'spearmanr' - sort by decreasing Spearman correlation. N spearmanrrpearsonr var_explained MAD_explainedzUnknown sorting metric: {})npisfinitescipystatsr r signvarrCOMEDMAD ValueErrorformatappendrangelenargsortabs) eigvalseigvecs phasing_track sort_metriccorrseigvecmaskcorriidxs T/var/www/html/software/conda/lib/python3.11/site-packages/cooltools/api/eigdecomp.py _phase_eigsr( s& E{6""R[%?%??  +"<"<;((t)-"5vd|DDx|t HHDD9@@MMNN N T3w<< 44WU1X&&3 j"&--(("3< G c ,tjdtj<ddk}jd|dzks||dzkrXtjdt |Dtjfdt |DfS|r-t | dz|D]}t jd|t j|\}} } } |rC|dkr=tj |dtj ||d d fd d |f|}|dz}d||d d f<d|d d |f<t j ||d \} } | tj tj | d zdd d d fz} | tj tj| d d d fz} |t| | ||\} } | | fS) a Compute compartment eigenvector on a dense cis matrix. Note that the amplitude of compartment eigenvectors is weighted by their corresponding eigenvalue Parameters ---------- A : 2D array balanced dense contact matrix n_eigs : int number of eigenvectors to compute phasing_track : 1D array, optional if provided, eigenvectors are flipped to achieve a positive correlation with `phasing_track`. ignore_diags : int the number of diagonals to ignore clip_percentile : float if >0 and <100, clip pixels with diagonal-normalized values higher than the specified percentile of matrix-wide values. sort_metric : str If provided, re-sort `eigenvecs` and `eigvals` in the order of decreasing correlation between phasing_track and eigenvector, using the specified measure of correlation. Possible values: 'pearsonr' - sort by decreasing Pearson correlation. 'var_explained' - sort by decreasing absolute amount of variation in `eigvecs` explained by `phasing_track` (i.e. R^2 * var(eigvec)) 'MAD_explained' - sort by decreasing absolute amount of Median Absolute Deviation from the median of `eigvecs` explained by `phasing_track` (i.e. COMED(eigvec, phasing_track) * MAD(eigvec)). 'spearmanr' - sort by decreasing Spearman correlation. This option is designed to report the most "biologically" informative eigenvectors first, and prevent eigenvector swapping caused by translocations. In reality, however, sometimes it shows poor performance and may lead to reporting of non-informative eigenvectors. Off by default. Returns ------- eigenvalues, eigenvectors .. note:: ALWAYS check your EVs by eye. The first one occasionally does not reflect the compartment structure, but instead describes chromosomal arms or translocation blowouts. raxisr*c&g|]}tjS)rnan.0r%s r' zcis_eig..ys444bf444r)chg|].}tjjdtjz/S)r)ronesshaper0)r2r%As r'r3zcis_eig..zs0JJJqbgagaj))BF2JJJr)g?dNTmask_zero_rowsr)rarrayrsumr6rrset_diagobserved_over_expectedclip percentileget_eigsqrtnansumrr() r7n_eigsr ignore_diagsclip_percentiler r#dOE_rrs ` r'cis_eigrK@s<d  AAr{1~~o 55a5==1 Dwqz\A%%%|a7G)G)G H44eFmm444 5 5 HJJJJE&MMJJJ K K  ) }q(,77 ) )A  aa ( ( ( (1!T::KB1aR?S00 WRBM"T111W+aaag*>PP Q Q#IBBuaaaxLBqqq4%xL'F4HHHGW rwryAA666774@@G rwrvg''400G &w TT G r)c<tj|||}||}||||k<|||<tj|d}|tj|ddk}tj||}|dk||kz} d|| ddf<d|dd| f<|S)Nrr,)rrAr=) r7 transmaskperc_top perc_bottomlimtdatamargmarg_nz min_cutoffdropmasks r'_filter_heatmaprVs -) h / /C iLEE%#+AiL 6!!   D26!!$$$q()Gw 44JqTJ./HAhkNAaaakN Hr)c|tj}tjtj|ddk}d|dd|f<d||ddf<t j|||S)Nrr,g|=r)astyperuint8rr=rfake_cis)r7cismaskss r' _fake_cisr]srnnRX&&G rvaa   !!U*AGAAAqDMGAqqqDM a!!! Hr)X@r8Fc:tj|}|jd|jdkrtd|jd}|ddkr6|d|kr*tjtj|dks"td|t|dd|dd}g} t|D]/\} \} } d|| | | | f<| | g| | z z0tj| } | dddf| dddfk} tj |ddk}|}tj ||}d||ddf<d|dd|f<t|| |z||}tj |ddk}|}tj ||}d||ddf<d|dd|f<t|| }tj|d}t|| }tj|d}tj||}||z |z }d||ddf<d|dd|f<tj||d \}}|tjtj |d zddddfz}|tjtj|dddfz}|t+||||\}}||fS) a  Compute compartmentalization eigenvectors on trans contact data Parameters ---------- A : 2D array balanced whole genome contact matrix partition : sequence of int bin offset of each contiguous region to treat separately (e.g., chromosomes or chromosome arms) n_eigs : int number of eigenvectors to compute; default = 3 perc_top : float (percentile) filter - clip trans blowout contacts above this cutoff; default = 99.95 perc_bottom : float (percentile) filter - remove bins with trans coverage below this cutoff; default=1 phasing_track : 1D array, optional if provided, eigenvectors are flipped to achieve a positive correlation with `phasing_track`. sort_metric : str If provided, re-sort `eigenvecs` and `eigvals` in the order of decreasing correlation between phasing_track and eigenvector, using the specified measure of correlation. Possible values: 'pearsonr' - sort by decreasing Pearson correlation. 'var_explained' - sort by decreasing absolute amount of variation in `eigvecs` explained by `phasing_track` (i.e. R^2 * var(eigvec)) 'MAD_explained' - sort by decreasing absolute amount of Median Absolute Deviation from the median of `eigvecs` explained by `phasing_track` (i.e. COMED(eigvec, phasing_track) * MAD(eigvec)). 'spearmanr' - sort by decreasing Spearman correlation. This option is designed to report the most "biologically" informative eigenvectors first, and prevent eigenvector swapping caused by translocations. In reality, however, sometimes it shows poor performance and may lead to reporting of non-informative eigenvectors. Off by default. Returns ------- eigenvalues, eigenvectors .. note:: ALWAYS check your EVs by eye. The first one occasionally does not reflect the compartment structure, but instead describes chromosomal arms or translocation blowouts. rr8zA is not symmetriczANot a valid partition. Must be a monotonic sequence from 0 to {}.Nr,Tr:r)rr<r6ralldiffrzip enumerateextendrD logical_andouterrVr]riterative_correction_symmetricmeanrBrCrr()r7 partitionrErNrOrr n_binsextentspart_idsni0i1is_trans is_bad_bin is_good_binis_validAbarOrrs r' trans_eigrwsTp  AwqzQWQZ-... WQZF! imv55"&ASASVWAW:X:X5 "F6NN   )CRC.)ABB-00GH )))) 8B"R%B,rBw((((x!!H4 HT111W$55H11%%%*J+K~##K==HAj!!!mAaaam8h.+FFA11%%%*J+K~##K==HAj!!!mAaaam !hYA/2215A!hYA/2215A 71X;  D TTAAj!!!mAaaam'6$GGGGW rwryAA666774@@G rwrvg''400G &w TT G r)weightgX@c .|t}n7 t|dd} n"#t$r} td| d} ~ wwxYw t d} n&#t$r} tdd| d} ~ wwxYw,dd d ndd} t|dd tj | | d gd } | j dd| j f} dtD} | D]}tj| |<|}dtD}|D]}tj||<fd}|||j}|D]W\}}}tj| |j}|j| j || f<tj||j}||j ||f<X|| fS)a Compute compartment eigenvector for a given cooler `clr` in a number of symmetric intra chromosomal regions defined in view_df (cis-regions), or for each chromosome. Note that the amplitude of compartment eigenvectors is weighted by their corresponding eigenvalue. Eigenvectors can be oriented by passing a binned `phasing_track` with the same resolution as the cooler. Parameters ---------- clr : cooler cooler object to fetch data from phasing_track : DataFrame binned track with the same resolution as cooler bins, the fourth column is used to phase the eigenvectors, flipping them to achieve a positive correlation. view_df : iterable or DataFrame, optional if provided, eigenvectors are calculated for the regions of the view only, otherwise chromosome-wide eigenvectors are computed, for chromosomes specified in phasing_track. n_eigs : int number of eigenvectors to compute clr_weight_name : str name of the column with balancing weights to be used. ignore_diags : int, optional the number of diagonals to ignore. Derived from cooler metadata if not specified. clip_percentile : float if >0 and <100, clip pixels with diagonal-normalized values higher than the specified percentile of matrix-wide values. sort_metric : str If provided, re-sort `eigenvecs` and `eigvals` in the order of decreasing correlation between phasing_track and eigenvector, using the specified measure of correlation. Possible values: 'pearsonr' - sort by decreasing Pearson correlation. 'var_explained' - sort by decreasing absolute amount of variation in `eigvecs` explained by `phasing_track` (i.e. R^2 * var(eigvec)) 'MAD_explained' - sort by decreasing absolute amount of Median Absolute Deviation from the median of `eigvecs` explained by `phasing_track` (i.e. COMED(eigvec, phasing_track) * MAD(eigvec)). 'spearmanr' - sort by decreasing Spearman correlation. This option is designed to report the most "biologically" informative eigenvectors first, and prevent eigenvector swapping caused by translocations. In reality, however, sometimes it shows poor performance and may lead to reporting of non-informative eigenvectors. Off by default. map : callable, optional Map functor implementation. Returns ------- eigvals, eigvec_table -> DataFrames with eigenvalues for each region and a table of eigenvectors filled in the `bins` table. .. note:: ALWAYS check your EVs by eye. The first one occasionally does not reflect the compartment structure, but instead describes chromosomal arms or translocation blowouts. Possible mitigations: employ `view_df` (e.g. arms) to avoid issues with chromosomal arms, consider blacklisting regions with translocations during balancing. NT) check_sorting raise_errorsz0view_df is not a valid viewframe or incompatibler{#provided cooler is not balanced or is missingzbins/rFrview_dfclr_weight_namemask_clr_bad_bins drop_track_na view_regionr)subsetr-cg|] }d|dz S)Er8r/r1s r'r3zeigs_cis..s"999ak!a%kk999r)cg|] }d|dz Seigvalr8r/r1s r'r3zeigs_cis..s%>>>1&q1u&&>>>r)c|dd} |} #tj |}|dj}nd}t | | \}}|||fS)a perform eigen decomposition for a given region assuming safety checks are done outside of this function. Parameters ---------- region: tuple-like tuple of the form (chroms,start,end,*) Returns ------- _region, eigvals, eigvecs -> ndarrays array of eigenvalues and an array eigenvectors Nr*balancevalue)rErFrrGr )matrixfetchbioframeselectvaluesrK)region_regionr7phasing_track_regionphasing_track_region_valuesrrrGclrrrFrErr s r'_eachzeigs_cis.._eachs!* JJJ / / 5 5g > >  $#+?='#J#J *>w*G*N ' '*. '" %5+#    ((r))rr Exceptionrr _load_attrsgetbinsrr assign_viewdropnaloccolumnsrrr0copyrrindexT)rrrrErrFrGr maprJer eigvec_tableeigvec_columnsev_col eigvals_tableeigval_columnseval_colrresultsr_eigvals_eigvecsr&s`` ````` r'eigs_cisr#sP"3'' X'"! AA  X X XOPPVW W X sO$ G G G  N/ N N N      1112266~qIII  88::aaa=D /  +"    'g66==m_[\=]]L#AAAt|O4L995==999N &&!v VLLNNM>>f >>>N"))"$& h!)!)!)!)!)!)!)!)!)!)!)Jc%((G(/::#8olG44:08 n,-omW55;19 #~-.. , &&s,/ AA  AA%% B/BBc \ t|d}n&#t$r}td|d|d}~wwxYwd} | t} nt d|Dt jfdjDt f}|d} |d} | | | | | f} | | } |+t|| |dd }|d j | | }nd}t| |f|||d |\}}| }t|D]\}}||d |dz<t!jt j|dt'|D}||fS)NTr|r}r~z2views are not currently implemented for eigs_transc:g|]}|Sr/)offset)r2chromrs r'r3zeigs_trans..s% ; ; ;5SZZ   ; ; ;r)rr`rrr)rErr rr8cg|] }d|dz Srr/r1s r'r3zeigs_trans..s%999a!!a%!!999r))datar)rrrrNotImplementedErrorrr_ chromnamesrrrrrrwrrdpd DataFrame atleast_2dr)rrrErjrr kwargsrJrrlohir7rphasing_track_valuesrrrr%r"s` r' eigs_transrs% sO$ G G G  N/ N N N    G"3''!"VWWWE ; ; ; ;CN ; ; ;S__ L  1B 2B ? ++BrE2b5L9A 88::be D /  +"     -W5rsb DDDDDDDDBBBBBBBB0000hUYUUUUp   "    nnnnf   m'm'm'm'd  A!A!A!A!A!A!r)