Now, compare this to a typical drug discovery program which synthesizes and tests only 1,000 molecules per year. Obviously, compute power is increasing and accelerating each year, which means that at some point in the future, Schrödinger will be able to explore 237 billion different compounds computationally per day. Schrödinger’s platform is able to explore 237 billion different compounds computationally every six months. What does this mean and why is this the right way to think about it?Īs you may know, all the chemicals in the universe are thought to be around 1060. The investment case with Schrödinger is all about the speed at which the company is covering ground in the drug discovery space. To look around the corner in Schrödinger, investors need to find the most important metric.
"Schrödinger is a fascinating company but market participants are truly struggling to find a way to think about the business model. info ( "Writing merged from glide poses file." ) reader = structure. add ( compound_id ) return True return False # It's preferable to have write the Glide poses out first: self. property = consensus_scores merged_writer. get ( compound_id ) if zscore is not None : st. property = zscore zscore = glide_zscores_by_compound. property = zscore zscore = canvas_zscores_by_compound. read_unique_field ( st, uniquefield ) if compound_id in keep_compounds and compound_id not in written_compounds : zscore = shape_zscores_by_compound. Shape results # FIXME current line #1 is supported written_compounds = set () def merge_compound ( st ): # Get compound code: compound_id = pipeutils. info ( "Keeping the best %d compounds" % len ( keep_compounds )) # No read the structures in this order (as available): # 1. sort ( reverse = True ) # Higher scores first keep_compounds = set () for score, compound in decorated_list : keep_compounds. info ( "Scored %d compounds" % len ( consensus_scores )) # Keep only the best number of compounds that was requested: decorated_list = decorated_list. calcConsensusScore ( scores, top_n ) consensus_scores = consensus_score self. append ( glide_zscore ) # Consensus score is ave of the best "top_n" (default=2) Z-scores: consensus_score = self. append ( canvas_zscore ) if glide_zscore is not None : scores. append ( shape_zscore ) if canvas_zscore is not None : scores. get ( compound_id ) # Use only scores from methods which produced hits, and only # from methods that produced 2 or more hits (otherwise the # z-score can't be calculated and is set to ): scores = if shape_zscore is not None : scores. get ( compound_id ) glide_zscore = glide_zscores_by_compound. get ( compound_id ) canvas_zscore = canvas_zscores_by_compound. """ labelfield = self labels = self inputs = for compound_id in all_compound_ids : shape_zscore = shape_zscores_by_compound. Raises a RuntimeError if there is a problem reading an input file or writing an output file.
def operate ( self ): """ Combine all the input files from all input sets into one set, optionally labelling each structure according to the set from which it originated. """ import sys from past.utils import old_div import numpy from schrodinger import structure from schrodinger.pipeline import pipeio from schrodinger.pipeline import pipeutils from schrodinger.pipeline import stage # Properties written out by the scoring methods: PHASE_SHAPE_SCORE_PROP = "r_phase_Shape_Sim" CANVAS_2D_SCORE_PROP = "r_canvas_FP_Similarity" GLIDE_SCORE_PROP = "r_i_docking_score" # Properties that are written by the DataFusionMergeStage: ZSCORE_FINGERPRINT_PROP = "r_datafusion_Z-score_Fingerprint" ZSCORE_SHAPE_PROP = "r_datafusion_Z-score_Shape" ZSCORE_DOCKING_PROP = "r_datafusion_Z-score_Docking" ZSCORE_AVERAGE_PROP = "r_datafusion_Z-score_Average" DataFusionMergeStage Stage for merging results in a Data Fusion workflow. CombineStage Stage for combining multiple structure PipeIO sets into one PipeIO set and labelling each ligand by the input set from which it came. # -*- coding: utf-8 -*- """ Core stages for Pipeline mechanism.
0 kommentar(er)
