2. ADMET Properties

2. ADMET Properties

ADMET Predictor has always been a powerful way to characterize molecular structures in terms of high-end descriptors and then use models built on those descriptors to predict the biologically relevant properties of the compound. Version 8.0 introduced a new, more flexible graphical user interface for viewing the corresponding predictions. You can now rearrange, hide or delete rows and columns; summarize data from multiple columns as star plots; link histograms and xy plots to row selections; and insert unlimited functional (e.g., log(x)) columns.

The models in ADMET Predictor fall into four broad groups that are linked to the spreadsheet's display tabs: physicochemical & biopharmaceutical (PhysChem); Transporters; Metabolism; and Toxicity. In the past, there was a one-to-one mapping of these model groups to the respective Module licenses. Since ADMET Predictor 8.0, some models are enabled by more than one Module license.

The PhysChem group includes models for ionization (pK_a), lipophilicity, solubility, permeability, transporters, pharmacokinetic parameters and antiviral inhibition.

The Metabolism group includes models for UDP-glucuronosyltransferase (UGT) specificities as well cytochrome P450 (CYP) sites of metabolism, kinetic parameters and inhibition patterns.

The Transporters group includes models to predict the likelihood that a compound is a substrate or inhibitor of one these key transporters along with kinetic parameters (i.e., Michaelis constants, or K_M's) for the uptake transporters.

The Toxicity group includes models for genotoxicity and ecotoxicity as well as toxicity specific to rodents and humans.

The Risks group is made up of ADMET Risk models, which very concisely summarize model information from the other groups. They highlight any predictions that fall outside the range of values observed for a reference set of drugs.

Unless otherwise indicated, models were developed and are distributed solely by Simulations Plus. In many cases, this is emphasized by the "S+" at the start of the model name.

The following list enumerates all 169-plus models offered in ADMET Predictor-X (by default), laid out as they are described in the documentation that follows.

ADMET Risk Models

1. ADMET_Risk and ADMET_Code: global ADMET Risk and ADMET Code summarizing all other ADMET Risk models. This tool, developed by Simulations Plus, identifies and communicates potential development liabilities in drug candidates. The criteria used are based on predictions for a refined reference subset drawn from the World Drug Index (WDI).

2. Absn_Risk and Absn_Code: ADMET Risk and ADMET Code for flagging potential oral absorption problems.

3. CYP_Risk and CYP_Code: ADMET Risk and ADMET Code for flagging potentially problematic metabolism.

4. MUT_Risk and MUT_Code: ADMET Risk and ADMET Code for mutagenicity in bacteria – a computational filter developed by Simulations Plus summarizing the output of MUT_* models that can serve as an in silico Ames test.

5. TOX_Risk and TOX_Code: ADMET Risk and ADMET Code for potential toxicity liabilities.

6. RuleOf5 and RuleOf5_Code: a Risk score and codes identical to Lipinski’s “Rule of Five” (1997).

Physicochemical and Biopharmaceutical (PhysChem) Properties

Dissociation Constants (pKa)

S+pKa uses several submodels to predict protic ionization microconstants for all identified ionizable groups in a molecule. The microconstants obtained are subjected to the microequilibria analysis to arrive at the macroscopic ionization constants (pK_a's) that can be observed in a standard titration experiment. This can be visualized directly, summarized in the spreadsheet in terms of macroscropic pK_a's (S+Acidic_pKa, S+Mixed_pKa, and S+Basic_pK_a), or as various flavors of titration profiles. In addition, the program calculates measures of individual ionizable group acidity (pK₅₀ values; Fraczkiewicz 2023) and presents these in the S+pK50 column.

Lipophilicity

1. MlogP: Moriguchi model of octanol-water partition coefficient, log P (Moriguchi 1992).

2. S+logP: octanol-water partition coefficient (log P).

3. S+logD: octanol-water distribution coefficient (log D) calculated from S+pKa and S+logP. This can be plotted as a function of pH.

4. LogHLC [atm*m³/mol]: logarithm of the air-water partition coefficient, a.k.a. Henry’s Law Constant (Simulations Plus model).

5. ChromLogD: chromatographic logD.

Solubility

1. S+S_Intrins [mg/mL]: intrinsic solubility.

2. S+pH_Satd: native pH calculated from S+pKa and S+S_Intrins.

3. S+Sw [mg/mL]: native water solubility calculated from S+pKa, S+S_Intrins, and S+pH_Satd.

4. SolFactor: salt solubility factor for S+ solubility models.

5. S+S_pH [mg/mL]: solubility at user specified pH calculated from S+pKa, S+S_Intrins and SolFactor. This can be plotted as a function of pH.

6. S+MP [deg C]: melting point (an important component of solubility).

7. DG_SelfSolv and DG_Hydration [kcal/mol]: Gibbs free energy of self solvation and hydration, respectively.

8. S+FaSSGF [mg/mL]: solubility in simulated fasted state gastric fluid.

9. S+FaSSIF [mg/mL]: solubility in simulated fasted state intestinal fluid.

10. S+FeSSIF [mg/mL]: solubility in simulated fed state intestinal fluid.

11. SupSatn [yes/no]: tendency of a compound to supersaturate in water.

Permeability

1. DiffCoef [cm2/s×10⁵]: molecular diffusion coefficient in water (Hayduk and Laudie 1974).

2. MolVol [cm³/mol]: molal volume at the normal boiling point (Schroeder 1929).

3. S+MDCK [cm/s×10⁷]: apparent MDCK Transwell permeability.

4. S+MDCK-LE [cm/s×10⁶]: Low efflux MDCK permeability.

5. S+PAMPA [High/Low]: PAMPA membrane permeability.

6. S+Peff [cm/s×10⁴]: human effective jejunal permeability.

7. Perm_Skin [cm/s×10⁷]: permeability through human skin.

8. Perm_Cornea [cm/s×10⁷]: permeability through rabbit cornea.

9. SimplePeff [cm/s ×10⁴]: simple linear model of human jejunal effective permeability.

10. BB_Filter: qualitative likelihood (High/Low) of crossing the blood-brain barrier.

11. LogBB: logarithm of the brain/blood partition coefficient.

Pharmacokinetic Properties

Distributional Properties

1. fup_hum: percent unbound to blood plasma proteins in human.

2. fup_rat: percent unbound to blood plasma proteins in rat.

3. fup_mou: percent unbound to blood plasma proteins in mouse.

4. fup_dog: percent unbound to blood plasma proteins in dog.

5. fup_mky: percent unbound to blood plasma proteins in monkey (cynomolgus and rhesus combined).

6. RBP_hum: blood-to-plasma concentration ratio in human.

7. RBP_rat: blood-to-plasma concentration ratio in rat.

8. RBP_mou: blood-to-plasma concentration ratio in mouse.

9. RBP_dog: blood-to-plasma concentration ratio in dog.

10. RBP_mky: blood-to-plasma concentration ratio in monkey.

11. fumic_hum: fraction unbound in human liver microsomes at 1 mg/mL microsomal protein.

12. fumic_rat: fraction unbound in rat liver microsomes at 1 mg/mL microsomal protein.

13. fumic_mou: fraction unbound in mouse liver microsomes at 1 mg/mL microsomal protein.

14. fumic_dog: fraction unbound in dog liver microsomes at 1 mg/mL microsomal protein.

15. fumic_mky: fraction unbound in monkey liver microsomes at 1 mg/mL microsomal protein.

16. S+fuhep: fraction unbound in human hepatocytes at 10⁶ cells/mL.

17. Vd [L/kg]: volume of distribution in humans at steady state.

Clearance Classification

1. ECCS_Class: Extended Clearance Classification System (ECCS) assignment.

2. S+CL_Mech: predicts whether metabolism, renal elimination or hepatic uptake will limit clearance.

3. S+CL_Metab: predicts whether or not metabolism will be critical to clearance.

4. S+CL_Renal: predicts whether or not renal elimination will be critical to clearance.

5. S+CL_Uptake: predicts whether or not hepatic uptake will be critical to clearance.

BCS/DCS Assignment

* Projects class assignments in the Biopharmaceutics Classification System *(BCS) and Developability Classification System (DCS) as a function of dose.

Transporters Models

1. Pgp_Substr: likelihood of P-glycoprotein efflux.

2. Pgp_Inh: likelihood of P-glycoprotein inhibition.

3. BCRP_Substr: likelihood of efflux on the Breast Cancer Resistance Protein (BCRP).

4. BCRP_Inh: likelihood of Breast Cancer Resistance Protein (BCRP) inhibition

5. OATP1B1_Substr: likelihood of the compound being a substrate of hepatic OATP1B3 transporter

6. OATP1B1_Inh: likelihood of the hepatic OATP1B1 transporter inhibition.

7. OATP1B1_Km: kinetic Michaelis-Menten Km constants for predicted substrates of OATP1B1 transporter

8. OATP1B3_Substr: likelihood of the compound being a substrate of hepatic OATP1B3 transporter

9. OATP1B3_Inh: likelihood of the hepatic OATP1B3 transporter inhibition.

10. OATP1B3_Km: kinetic Michaelis-Menten Km constants for predicted substrates of OATP1B3 transporter

11. BSEP_Inh: likelihood of Bile Salt Export Pump (BSEP) transporter inhibition.

12. BSEP_IC50: IC₅₀ for inhibition of BSEP transporter with taurocholate as a probe substrate

13. OAT1_Substr: likelihood of the compound being a substrate of renal OAT1 transporter

14. OAT1_Inh: likelihood of the renal OAT1 transporter inhibition.

15. OAT1_Km: kinetic Michaelis-Menten Km constants for predicted substrates of OAT1 transporter

16. OAT3_Substr: likelihood of the compound being a substrate of renal OAT3 transporter

17. OAT3_Inh: likelihood of the renal OAT3 transporter inhibition.

18. OAT3_Km: kinetic Michaelis-Menten Km constants for predicted substrates of OAT3 transporter

19. OCT1_Substr: likelihood of the compound being a substrate of hepatic OCT1 transporter

20. OCT1_Inh: likelihood of the hepatic Organic Cation Transporter 2 (OCT1) inhibition.

21. OCT1_Km: kinetic Michaelis-Menten Km constants for predicted substrates of OCT1 transporter

22. OCT2_Substr: likelihood of the compound being a substrate of renal OCT2 transporter

23. OCT2_Inh: likelihood of the renal OCT2 transporter inhibition.

24. OCT2_Km: kinetic Michaelis-Menten Km constants for predicted substrates of OCT2 transporter

Metabolism Models

Phase I: Oxidation

1. CYPxxx_Substr: substrate classification models for nine human CYPs: 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1 and 3A4.

2. CYPxxx_Inh: general inhibitor classification models for CYPs 1A2, 2C9, 2C19, 2D6 and 3A4.

3. CYPxxx_Km [μM]: estimated Michaelis-Menten K_m constants for predicted sites of metabolism by five major human CYPs: 1A2, 2C9, 2C19, 2D6 and 3A4.

4. CYPxxx_Vmax [nmol/min/nmol enzyme]: estimated Michaelis-Menten V_max constants for predicted sites of metabolism by CYPs 1A2, 2C9, 2C19, 2D6 and 3A4.

5. CYPxxx_CLint [μL/min/mg HLM protein]: estimated intrinsic clearance for predicted sites of metabolism by recombinant CYPs 1A2, 2C9, 2C19, 2D6 and 3A4.

6. CYPSum_CLint: sum of the above CYPxxx_CLint predictions.

7. CYPxxx_Ki [μM]: inhibition K_i constants for predicted sites of metabolism by seven major human CYPs: 1A2, 2A6, 2C8, 2C9, 2C19, 2D6 and 3A4.

8. CYP3A4_HLM_Km [μM]: kinetic Michaelis-Menten K_m constants for predicted sites of CYP 3A4-mediated metabolism by human liver microsomes (HLMs).

9. CYP3A4_HLM_Vmax [nmol/min/mg HLM protein]: kinetic Michaelis-Menten Vmax constants for predicted sites of CYP 3A4-mediated metabolism by HLMs.

10. CYP3A4_HLM_CLint [μL/min/mg HLM protein]: intrinsic clearance constants for predicted sites of CYP 3A4 mediated metabolism in HLMs.

11. AOX_Substr. substrate classification model for Aldehyde Oxidase.

Intrinsic clearance

1. CYP_HLM_CLint [μL/min/mg HLM protein]: overall in vitro (unbound) intrinsic clearance in Human Liver Microsomes.

2. CYP_RLM_CLint [μL/min/mg RLM protein]: overall in vitro (unbound) intrinsic clearance in Rat Liver Microsomes.

3. CYP_MLM_CLint [μL/min/mg MLM protein]: overall in vitro (unbound) intrinsic clearance in Mouse Liver Microsomes.

4. CYP_DLM_CLint [μL/min/mg MLM protein]: overall in vitro (unbound) intrinsic clearance in Dog Liver Microsomes.

5. CYP_MKYLM_CLint [μL/min/mg MLM protein]: overall in vitro (unbound) intrinsic clearance in Monkey Liver Microsomes (cynomolgus and rhesus).

6. HEP_hCLint [μL/min/10⁶ cells]: overall in vitro (unbound) intrinsic clearance in human hepatocytes.

7. HEP_rCLint [μL/min/10⁶ cells]: overall in vitro (unbound) intrinsic clearance in rat hepatocytes.

8. HEP_mCLint [μL/min/10⁶ cells]: overall in vitro (unbound) intrinsic clearance in mouse hepatocytes.

9. HEP_dCLint [μL/min/10⁶ cells]: overall in vitro (unbound) intrinsic clearance in dog hepatocytes.

10. HEP_mkyCLint [μL/min/10⁶ cells]: overall in vitro (unbound) intrinsic clearance in monkey hepatocytes (cynomolgus and rhesus).

Phase II: Glucuronidation

1. UGTxxx: nine substrate classification models for glucuronidation by UGT1A1, UGT1A3, UGT1A4, UGT1A6 , UGT1A8, UGT1A9, UGT1A10, UGT2B7 and UGT2B15 (nine models; Enslein Research and Simulations Plus).

Predicting Metabolites

1. CYPxxx_Sites: sites of metabolism models for oxidation by nine human CYPs 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1 and 3A4.

2. AOX_Sites: sites of aldehyde oxidase (AOX) mediated oxidation.

Note that while we do not have site models for UGTs or esterases, you can still predict metabolites for these enzymes using either ADMET Predictor or MedChem Designer.

Toxicity Models

Reproductive and Genetic Toxicity

1. Repro_Tox: qualitative estimation of reproductive / developmental toxicity.

2. Chrom_Aberr: qualitative estimation of the risk of triggering the mutagenic chromosomal aberrations.

3. MUT_xxx: classification model for the mutagenicity of pure compounds in S. typhimurium strains TA98 (98); TA_100 (100); TA1535 (1535); and TA97 or TA1537 (97+1537).

4. MUT_mxxx: classification model for the mutagenicity of compounds after rat liver microsomal activation in S. typhimurium strains TA98 (98); TA_100 (100); TA1535 (1535); and TA97 or TA1537 (97+1537).

5. MUT_102+wp2: classification model for the mutagenicity of pure compounds in S. typhimurium strain TA102 or E. coli strain WP2 uvrA.

6. MUT_m102+wp2: classification model for mutagenicity of compounds after rat liver microsomal activation in S. typhimurium strain TA102 or E. coli strain WP2 uvrA.

7. MUT_NIHS: classification model for predicting the mutagenicity of compounds based on data compiled by the Japanese National Institute of Health Sciences (NIHS) in 2017.

Rodent Toxicity

1. Sens_Skin: qualitative assessment of allergenic skin sensitization in mouse.

2. Sens_Resp: qualitative assessment of allergenic respiratory sensitization in rat.

3. Rat_Acute [mg/kg]: LD₅₀ for lethal rat acute toxicity by any mechanism.

4. Rat_TD50 [mg/kg/day]: median chronic toxic dose (TD₅₀), which is defined as the oral dose of a compound required to induce tumors in 50 percent of a rat population after exposure over a standard lifetime.

5. Mouse_TD50 [mg/kg/day]: chronic TD₅₀ for mouse.

Human Toxicity

1. hERG_Filter: qualitative estimation of the likelihood of the hERG potassium channel inhibition in human.

2. hERG_pIC50 [mol/L]: pIC₅₀ as a measure of affinity towards hERG K+ channel and potential for cardiac toxicity.

3. Ser_AlkPhos: human liver adverse effect as the likelihood of causing elevation in the levels of Alkaline Phosphatase enzyme.

4. Ser_GGT: human liver adverse effect as the likelihood of causing elevation in the levels of GGT enzyme.

5. Ser_LDH: human liver adverse effect as the likelihood of causing elevation in the levels of LDH enzyme.

6. Ser_AST: human liver adverse effect as the likelihood of causing elevation in the levels of SGOT enzyme.

7. Ser_ALT: human liver adverse effect as the likelihood of causing elevation in the levels of SGPT enzyme.

8. MaxRTD [mg/kg/day]: qualitative assessment of the Maximum Recommended Therapeutic Dose administered as an oral dose.

9. PLipidosis: qualitative estimation of potential for causing phospholipidosis.

Ecotoxicity

1. Bioconcn: [C_fish/C_water]: bioconcentration factor - partition coefficient between fish tissues and environmental water at steady state.

2. Biodegradn: likelihood of biodegradation in the environment expressed as relative biological oxygen demand.

3. Th_pyr_pIGC50 [mM]: pIGC₅₀ for Tetrahymena pyriformis growth inhibition toxicity.

4. Daphnia_LC50 [mg/L]: LC₅₀ for Daphnia magna (water flea) lethal toxicity after 48 hours of exposure.

5. Minnow_LC50 [mg/L] (LC₅₀ for fathead minnow lethal toxicity after 96 hours of exposure.

6. Estro_Filter: classification model for predicting antiestrogen activity in rats.

7. Estro_RBA [100%*IC₅₀(estrogen)/IC₅₀(compound]: a quantitative measure of binding to rat androgen receptor. Note: this estimate is valid only for compounds classified as ‘Toxic’ by Estro_Filter.

8. Andro_Filter: classification model for predicting antiandrogen activity in rats.

9. Andro_RBA [100%*IC₅₀(androgen)/IC₅₀(compound)]: a quantitative measure of binding to rat androgen receptor binding. Note: this estimate is only for compounds classified as ‘Toxic’ by Andro_Filter.